Birkinshaw 1998

Mate choice in Pros
(Horn) (Coleoptera:Bostrichidae):
The role of male-produced aggregation pheromone
Lucy A. Birkinshaw
University of Leicester
Thesis submitted for the degree of

Doctor of Philosophy
JANUARY 1998
UMI Number: U106171
All rights reserved
INFORMATION TO ALL USERS

The quality of this reproduction is dependent upon the quality of the copy submitted.
In the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if material had to be removed,
a note will indicate the deletion.
Dissertation Publishing
UMI U106171
Published by ProQuest LLC 2013. Copyright in the Dissertation held by the Author.
Microform Edition © ProQuest LLC.
All rights reserved. This work is protected against
unauthorized copying under Title 17, United States Code.
ProQuest LLC
789 East Eisenhower Parkway
P.O. Box 1346
Ann Arbor, Ml 48106-1346
ACKNOWLEDGMENTS
I thank my supervisor, Professor Robert ‘thrasher’ Smith, for a great three

years of constructive criticism, discussion, editing, and taking my ankles out at unihoc.
Similarly, I thank my second supervisor, Dr. Rick Hodges. Without Rick’s support I know
the work performed in Ghana would not have got past Rob’s initial reaction of, ‘No you
can’t go to Africa!’.
The field work would also not have been possible without the support of the
Post Harvest Management Division, Ministry of Food and Agriculture in Ho. I thank Mr.
Kwaku Nicol, Mr. Sam Addo and the staff at the ministry for their help. We thank the
Forestry Department in Ho for permission to work on their land.
This project was funded by the Natural Resources Institute in Chatham, Kent,
UK. I thank Professor David Hall and Dudley Farman (NRI) for answering my questions
on the chemistry of pheromone signalling. Dudley Farman also kindly prepared and
analysed filters used for pheromone collection in Ghana.
I thank Mick Ward for technical support in Leicester. Ted Gaten, Stans and
Ewen taught me all I know about computers. Bev Sherbon and Dr. Muriel Walker helped
me take, develop and print pictures of my dissections.
Sterilization of insects was made possible by the kind permission and

assistance of Dr. David Bonnett at Leicester Royal Infirmary (chapter 7). Initial questions
about sperm competition were readily answered by Dr. Paul Eady. Later on Dr. Matt Gage
took on the role of ‘advisor on sperm’ and also taught me the subtle art of sperm counting.
I must thank those who have recently gone before me, Dr. Manuel Vazquez-
Arista, Dr. Henry Fadamiro and Dr. Dagmar Scholz. All of these LGB Doctors have
readily given advice/ reprints and best wishes.
I thank my aged ancestors, damned average genetic inheritance! Life in

Leicester has always been fun. Cheers in particular to Chris, Matt, Sharon and Boz; Jens
and my mate Vivi; Stans, Ems, Mike, Noosie, Pips, Helen, Karen and David. My final
word is, ‘beetles are NOT just bird food!’.
ABSTRACT
Lucy A. Birkinshaw 1998: Mate choice in Prostephanus truncatus (Horn)

(Coleoptera: Bostrichidae): The role of male-produced aggregation pheromone.
Prostephanus truncatus (Horn) (Coleoptera: Bostrichidae) is a destructive pest

of stored maize and cassava that has recently been accidentally introduced into tropical
Africa. Males produce an aggregation pheromone when on food, that attracts dispersing
males and females. P. truncatus aggregation pheromone is being used to monitor the
spread of P. truncatus (Larger Grain Borer) across Africa. The biological function of this
pheromone is controversial. This thesis investigates the role of aggregation pheromone in
mate choice in P. truncatus.
The literature on Coleopteran aggregation pheromones was reviewed, with

particular reference to the possible adaptive functions of aggregation pheromones.
Variation in Prostephanus truncatus aggregation-pheromone signalling was

detected. Conspecifics can detect these differences and are preferentially attracted to some
males more than others. Both males and females ‘agree’ which males are most attractive
(shown in a laboratory bioassay and in trapping experiments in the field). Females also
discriminate between potential mates on contact by a stylised pushing behaviour. Some
males consistently secure more matings than others when two males are presented at once
to a female. Discrimination between males mediated on contact through pushing is not
influenced by the male’s aggregation pheromone signal (both natural variation and
manipulation of the pheromone signal were studied).
Observation of adult beetles in an artificial host sandwiched between two glass

plates revealed that males and females pair up, and cohabit within the same tunnel system.
Pairs mate multiply (up to 20 times per 12 hours) and dissection of recently mated females
revealed that males deliver an oversized ejaculate (approx. 50 000 sperm) as an oval
spermatophore. Male investment in ejaculate was not found to be influenced by male
crowding or the presence of Female Factor (an involatile pheromone produced by females,
which can trigger aggregation pheromone shut down in males).
These results are discussed in the context of sexual-selection theory and also
with reference to the Integrated Pest Management of this insect.
CONTENTS
CHAPTER 1: INTRO DUCTIO N............................................................................1
1.1 THE INSECT..............................................................................................................................1

1.1.1 Pest status in Africa............................................................................................................. 1
1.1.2 Life history........................................................................................................................... 1
1.1.3 Host range............................................................................................................................ 2
1.1.4 Research emphasis to date..................................................................................................3
1.2 STARTING HYPOTHESIS.....................................................................................................4
1.3 SEXUAL SELECTION............................................................................................................ 5

1.3.1 Low sexual dimorphism of LGBexternal morphology..................................................5
1.4 SUMMARY OF CHAPTERS................................................................................................ 6
1.5 GENERAL MATERIALS ANDMETHODS.........................................................................7

1.5.1 Source of insects..................................................................................................................7
1.5.2 Source of plant material used as insect hosts.................................................................. 7
1.5.3 Insect cultures...................................................................................................................... 7
1.5.4 Experimental conditions.................................................................................................... 8
1.5.5 Insect handling.................................................................................................................... 8
1.5.6 Sexing Prostephanus truncatus........................................................................................ 8
1.5.7 Video equipm ent................................................................................................................ 9
1.5.8 Statistical analysis............................................................................................................. —9
CHAPTER 2: LITERATURE REVIEW OF COLEOPTERAN

AG G REG ATIO N PHERO M O NES..................................................................... 1 0
2.1: DEFINITION........................................................................................................................ 10
2.2: OCCURRENCE.................................................................................................................... 10
2.3: NATURE.................................................................................................................................11
2.4: BIOLOGICALFEATURES...................................................................................................12
Pheromone produced only when signaller is on the host:....................................................12
Response to pheromone is often increased by confirmation of the presence of food
via perception of food volatiles:...............................................................................................
Produced by species with a relatively long lived adult stage:............................................... 14
The food source is capable of supporting many individuals:..............................................14
The identity of the signaller is sex specific:............................................................................ 14
The pattern of response is sex specific:.................................................................................... 15
Production and/or response to pheromone often decreases with mating experience:..... 17
2.5: ANTI-AGGREGATION PHEROMONES........................................................................ 18
2.6: NON ADAPTIVE EXPLANATIONS................................................................................ 19
2.7: ADAPTIVE EXPLANATIONS........................................................................................... 19
2.8: COSTS OF SIGNALLING.................................................................................................20
2.9: BENEFITS OF SIGNALLING...........................................................................................20
2.10: COSTS AND BENEFITS OF RESPONDING TO PHEROMONE.........................23
2.11: HYPOTHESES DISCUSSED IN TERMS OF THE BIOLOGICAL FEATURES

OF AGGREGATION PHEROMONES...................................................................................... 23
Pheromone produced only when signaller is on the host:................................................... 23
Species are long lived as an adult:............................................................................................23
The food source is capable of supporting many individuals:..............................................23
The identity of the signaller is sex specific:............................................................................24
The pattern of response is sex specific:....................................................................................24
Production and/or response to pheromone often decreases with mating experience:..... 25
A nti-aggregants:.......................................................................................................................... 25
2.12: FUTURE DIRECTIONS.................................................................................................... 26
2.13: CONCLUSIONS.................................................................................................................. 2 7
CHAPTER 3: BASIC PATTERNS OF AGGREGATION-PHEROMONE

SIG NALLING AND RESPONSE 28
3.1 INTRODUCTION....................................................................................................................28
3.2 BIOASSAY DESIGN............................................................................................................. 3 0

3.2.1 Details of the apparatus.................................................................................................... 31
3.2.2 Procedure during the assay............................................................................................. 31
3.3 MATERIALS AND M ETHODS......................................................................................... 33

3.3.1 Influence of time of d a y ..................................................................................................33
3.3.2 Influence of using flour instead of whole grain as food, onthe signaller................ 33
3.3.3 Timing of shut down of signal induced by removal of food.................................... 34
3.3.4 Timing of onset of signalling after placing on grain.................................................. 3 4
3.3.5 Influence of plant host on male-aggregation-pheromone signal............................. 35
3.3.6 Timing of shut down of pheromone signal induced by Female Factor...................36
3.4 RESULTS...................................................................................................................................
3.4.1 Influence of time of d a y ................................................................................................. 37
3.4.2 Influence of using flour instead of whole grain as food, on the signaller.............. 38
3.4.3 Timing of shut down of signal induced by removal of food.................................... 38
3.4.4 Timing of onset of signalling after placing on grain.................................................. 39
3.4.5 Influence of plant host on male-aggregation-pheromone signal............................. 40
3.4.6 Timing of shut down of pheromone signal induced by Female Factor...................42
3.5 DISCUSSION 42
CHAPTER 4: INTER-MALE VARIATION IN PHEROMONE SIGNAL (I

BIOASSAY AND II CHEMICAL ANALYSIS) 45
4.1 INTRODUCTION 45
4.2 A IM S 47
4.3 MATERIALS AND METHODS 47

4.3.1 Experiment to determine if variation in male signal was detectable using the
bioassay......................................................................................................................................... 47
4.3.2 Experiment to determine if males and females ‘agree’ which of a pair of males
is more attractive.......................................................................................................................... 49
4.3.3 Experiment to determine if the response of females to aggregation pheromone
can be depressed by increasing their exposure to the signal prior to testing.....................50
4.3.4 Experiment to determine if variation exists between males of the same source
and age, and if so, whether the pattern of variation is stable over time............................... 51
4.3.5 Experiment to correlate bioassay response to chemical assay, for pheromone
plumes from single males...........................................................................................................51
4.4 RESULTS................................................................................................................................. 5 3
4.4.1 Experiment to determine if variation in male signal was detectable using the
bioassay......................................................................................................................................... 53
4.4.2 Experiment to determine if males and females ‘agree’ which of a pair o f males
is more attractive.......................................................................................................................... 54
4.4.3 Experiment to determine if the response of females to aggregation pheromone
can be depressed by increasing their exposure to the signal prior to testing.....................56
4.4.4 Experiment to determine if variation exists between males of the same source
and age, and if so, whether the pattern of variation is stable over time............................... 56
4.4.5 Experiment to correlate bioassay response to chemical assay of pheromone plumes
from single males.........................................................................................................................58
v
4.5 DISCUSSION 60
CHAPTER 5: INTER-MALE VARIATION IN PHEROMONE SIGNAL (III:

FIELD TRIALS)...................................................................................................... 6 3
5.1 INTRODUCTION...................................................................................................................63
5.2 A IM S........................................................................................................................................ 64
5.3 MATERIALS AND METHODS........................................................................................ 65

5.3.1 Trapping trials:................................................................................................................. 65
5.3.2 Pheromone sam ples:........................................................................................................69
5.4 RESULTS.................................................................................................................................71
5.4.1 Trapping trials.................................................................................................................. 71
5.4.2 Pheromone sam ples.........................................................................................................75
5.4.3 Additional observations................................................................................................... 75
5.5 DISCUSSION......................................................................................................................... 7 5
CHAPTER 6: MATE CHOICE ON CONTACT............................................. 7 9
6.1 INTRODUCTION...................................................................................................................7 9
6.2 A IM S.........................................................................................................................................81
6.3 DESCRIPTION OF COURTSHIP BEHAVIOUR IN PR O STEPH AN U S

TRU NC ATU S WHEN OUTSIDE THE PLANT HOST...................................................... 82
6.4 MATERIALS AND METHODS........................................................................................ 83

6.4.1 Experiment to test if Prostephanus truncatus is capable of mating within its plant
host................................................................................................................................................ 83
6.4.2 Observations of Prostephanus truncatus behaviour in tunnels within an artificial
host (recorded using time-lapse photography)......................................................................83
6.4.3 Method for recording pushing behaviour.....................................................................84
6.4.4 Preliminary investigation of courtship behaviour.......................................................84
6.4.5 Investigation into the relationship between pheromone attractiveness and success in
mating trials................................................................................................................................. 86
6.4.6 The influence of manipulating pheromone signal on courtship behaviour...........87
6.5 RESULTS................................................................................................................................. 88
6.5.1 Experiment to test if Prostephanus truncatus is capable of mating within its plant
host................................................................................................................................................ 88
6.5.2 Observations of Prostephanus truncatus behaviour in tunnels within an artificial
host (recorded using time-lapse photography)..................................................................... 88
6.5.4 Preliminary investigation of courtship behaviour...................................................... 90
6.5.5. Investigation into the relationship between pheromone attractiveness and success
in mating trials.............................................................................................................................. 91
6.5.6 The influence of manipulating pheromone signal on courtship behaviour.......... 97
6.6 DISCUSSION......................................................................................................................... 99
CHAPTER 7: SPERM C O M P E TITIO N ........................................................ 1 0 5
7.1 INTRODUCTION................................................................................................................ 105
7.2 A IM S...................................................................................................................................... 107
7.3 METHODS............................................................................................................................. 108

7.3.1 Dissection of reproductive organs................................................................................108
7.3.2 Sterilization of males for calculation of P2.................................................................108
7.3.3 Determination of sperm number per ejaculate for two consecutive matings of
males kept in different sociosexual environments................................................................108
7.4 RESULTS............................................................................................................................... I l l
7.4.1 Dissection of reproductive organs................................................................................I l l
7.4.2 Sterilization of males for calculation of P2.................................................................114
7.4.3 Determination of sperm number per ejaculate for two consecutive matings of
males kept in different sociosexual environments................................................................114
7.5 DISCUSSION....................................................................................................................... 122

7 .5 .1 122
CHAPTER 8: DISCU SSIO N............................................................................1 2 7
8.1 Evolution of the aggregationpheromone signal/ response:...................................... 127
8.2 Alternative mating tactics:..............................................................................................131
8.3 The limitations of an evolutionaryapproach:.............................................................. 132
8.4 Practical applications of this work:.................................................................................132
8.5 Suggestions for future work:...........................................................................................133
R E F E R E N C E S ......................................................................................................1 3 4
i;
Fig. 1: Adult Prostephanus truncatus (length 2.5-4.5mm), and larva

(reproduced from GASGA/ CTA technical leaflet N o.l (1993)).
Chapter I: Introduction
CHAPTER Is INTRODUCTION
1.1 THE INSECT

1.1.1 Pest status in Africa
Prostephanus truncatus (Horn) (Coleoptera: Bostrichidae) is a destructive pest
of stored maize and cassava that has recently been accidentally introduced into tropical
Africa. Its presence was first officially confirmed in the Tabora region of Tanzania in the
early 1980’s (Cross, 1985). Since this time it has spread across six countries in the east of
Africa and what is thought to be a second introduction (Vazquez-Arista, 1997), now
covering at least seven West African countries has radiated from Togo (Scholz, 1997). P.
truncatus has most likely arrived in totally new locations by hitching a ride in grain
shipments. Local ranges are also spreading with time, presumably via insect dispersal by
flight (Fadamiro, 1995). Prostephanus truncatus originates from Mexico and Central
America where it is known only as a pest of stored maize.
The biology and pest status of Prostephanus truncatus are reviewed in detail in
Markham, Wright and Rios Ibarra (1991) and Hodges (1986; 1994). Recent reviews are
included in Fadamiro (1995) and Scholz (1997), therefore only the most relevant details are
given here. Larger Grain Borer is the most widely used common name for Prostephanus
truncatus in the scientific literature, and LGB will be the abbreviated form used throughout
this thesis.
Most of the damage caused by LGB is from the tunnelling activity of adults and
larvae rather than actual consumption of the host. At its worst P.truncatus reduced yields
by up to 34% weight loss in a 3-6 month storage period (Hodges et al., 1983, quoted in
Hodges, 1986). Patches of high level damage are in part a consequence of the aggregation
behaviour of Prostephanus truncatus. It was noticed early on that, although one store may
become seriously infested, neighbouring stores, presumably equally suitable as hosts,
would remain untouched. This phenomenon has since been attributed to a system of
aggregation pheromone communication in LGB. Males that locate a suitable host produce a
volatile chemical signal that is attractive to both sexes. Aggregation on food hosts has most
certainly been key to the creation of conflict between the interests of man and insect and will
be the focus of this study.
1.1.2 Life history

LGB is a fairly small beetle, typical of many stored product pests (approx.
adult length= 3-5m m , w idth= 1-1.5mm, fresh w eight= 2.5-4.5m g). Stored-product pests
com m only fall into one of two life-history strategies: short adult life where many small eggs
are produced in a short period o f time; long adult life with an extended reproductive period,
1
where a few fairly large offspring are produced at a constant rate (Burkholder and Ma,
1985). LGB falls into the second category (Li, 1988).
If a suitable host is available a female can consistently lay an average of four

eggs per day pausing only to construct new tunnels (Nyakunga, 1982, quoted in Hodges,
1986). Mean lifetime-fecundity estimates vary but a figure of 430 eggs is given by Bell and
Watters (1982) (quoted in Fadamiro, 1995 for LGB cultured on maize). Egg production
peaks at about three weeks after emergence and declines slowly from then on until death
(see Bell and Watters, 1982, quoted in Fadamiro, 1995). Both males and females will mate
multiply in the laboratory (pers. obs.) and females require at least three matings spaced
approximately one month apart to maximise the number of viable offspring she can produce
(Li, 1988).
When LGB is reared on maize under laboratory conditions of 32°C and 80%
r.h., small clutches of about eight eggs develop at the end of blind ending tunnels bored
into the host and hatch into cream coloured larvae after about five days (Li, 1988). Larvae
pass through three instars (possibly five instars see Ramirez, 1990, quoted in Markham et
al., 1991) and metamorphose into pupae after about three weeks. Pupae are often
surrounded by a thin-walled, closed cylindrical case in part constructed of flour. Beetles
remain as pupae for about four to five days, then hatch into lightly sclerotised adults. These
newly hatched adults are relatively inactive for their first few days, reach sexual maturity
after 4-5 days and are reproductively active for the remainder of their life span. LGB adults
live for about three to four months under ideal laboratory conditions.
1.1.3 Host range

Not surprisingly, diet greatly influences many life-history characters of this
beetle. This is an important consideration since LGB is not constrained to one host and has
been found breeding in plant species of different plant families. The life history described
above is based on studies of LGB on maize, currently its main host in the stored-product
environment. It is likely that LGB’s ancestors were woodborers as Bostrichids are
generally wood boring in habit (Fisher, 1950). Indeed, LGB has been reared in the
laboratory on the wood taken from trees of the Anacardiaceae, Burseraceae, Euphorbiaceae
and Leguminosae (Helbig and Schulz, 1994; Nang’ayo et a l, 1993, quoted in Fadamiro,
1995).
One mystery is the difficulty in locating LGB in wood in the field. LGB has yet
to be found living and breeding in wood in the wild except when it tunnels into grain-store
structures. A considerable population away from cultivation has been demonstrated (Rees et
al., 1990, in woodland areas of Mexico; and Nang’ayo et a l ., 1993, in Tsavo national
park, Kenya - all quoted in Fadamiro, 1995). Traps baited with artificially produced
aggregation pheromone were found to attract LGB just minutes after they were set up in
2
Chapter /: Introduction
areas many kilometres away from the nearest known source of maize (Hodges pers. comm,
in a bush area in Kenya.). Large aggregations of beetles have never been found naturally
occurring in a wood host despite considerable effort to identify such populations (Hodges
pers. comm.). This contrasts with the large aggregations of beetles that may be found in
food stores. It may be the case that large aggregations are a feature only of the stored-
product environment where food only becomes limiting only for very large populations, or
perhaps the location of aggregations in wood remains to be discovered. Ramirez et a l .,
(1991), quoted in Markham et al., (1991), suggested that LGB inhabits the drying out ends
of dead branches and twigs as a host, possibly those recently killed by the action of
Cerambycid beetles. This is compatible with the theory that LGB is particularly well suited
to outcompeting other species in relatively dry conditions (Howard, 1983; Markham et al,
1991).
1.1.4 Research em phasis to date

Much research effort has addressed the aim of damage limitation of this pest.
The main challenge has been to devise solutions that are practical and acceptable to small
scale farmers, and to provide this information through a network of extension workers. It is
unlikely that Prostephanus truncatus will be eliminated from Africa, but the devastation of
its initial introduction need not be repeated. A truly Integrated Pest Management approach
has been employed against LGB. Maize variety, storage practices, use of pesticides, use of
inert admixes, and biological control, are all factors that have been considered in an effort to
limit LGB damage. For example, Teretriosoma nigrescens (Histeridae) is a predatory
beetle found naturally in the American continent, which feeds on LGB larvae. Since it was
first isolated in infested maize at Reading University in the late 1970’s (Howard, 1983), it
has been studied as a potential biocontrol agent. Teretriosoma nigrescens is attracted to the
aggregation pheromone of LGB and uses this signal to locate its prey. T. nigrescens has
now been released in both East and West Africa as part of a biocontrol program against
LGB (Boeye etal., 1994).
Two components of LGB aggregation pheromone have been identified and

given the trivial names of Trunc-call 1 and Trunc-call 2, (hereafter referred to as T1 and
T2). T1 is l-methylethyl(E)-2-methyl-2-pentenoate and T2 is l-methylethyl(E,E)-2,4-
dimethyl-2,4-heptadienoate. These components can be manufactured by man and are used
to bait pheromone traps that have been successfully used to study LGB biology and to
monitor the spread of LGB (Hodges, 1986). Initially pheromone-baited traps were
employed in store to provide an early warning of LGB infestation. Unfortunately, once
LGB tunnel into a suitable host, they become relatively unresponsive to aggregation
pheromone. Regrettably these traps tended to attract the dispersing population around the
store instead of the population already within store, and thus they probably promoted
infestation of the store (Pike, 1993). It is now recommended that pheromone traps for
3
monitoring be placed at least 100m away from any stores (Hodges and Pike, 1995). The
occurrence and spread of T.nigrescens has been monitored using these pheromone traps
(Markham et al., 1994).
Host-plant volatiles are often used to increase the efficiency of aggregation-

pheromone traps for other pest species, as they can also be attractive and indeed act
synergistically with pheromone lures (see chapter 2). So far this has not been found to be
the case in LGB. As yet, no strong directional response to plant volatiles from a distance
has been observed (Fadamiro, 1995; Scholz et al., 1997b) although some short range
arrestant activity has been demonstrated for maize and cassava (Pike et al., 1994). For
general reviews of the use of pheromones as tools in stored product pest management see
Phillips (1994), and Chambers (1990).
Semiochemicals have so far fulfilled only part of the initial vision for their
application as a tool in pest management (Silverstein, 1990). It is fair to say that the main
successes have been realised in monitoring and the other main potential application, mating
disruption, has been relatively ineffective. Lack of understanding of the natural functioning
of semiochemicals has been highlighted as the main shortcoming of programmes utilising
these biological signals. The work presented here is a study of the reproductive biology of
Prostephanus truncatus with particular reference to the possible evolutionary advantages of
aggregation pheromone signalling. Aggregation pheromone signals (those that attract both
sexes) are a common feature of pest species so insights gained in this study are potentially
applicable to other systems of economic importance.
1.2 STARTING HYPOTHESIS

This thesis starts with the hypothesis that males signal to attract mates. A
corollary is that other males attracted are exploiting the signaller and represent a cost of
signalling through increased competition for mates. The finding that only males signal
using this pheromone was influential in the formulation of the hypothesis that Prostephanus
truncatus aggregation pheromone serves a sexual function, first formally proposed in print
by Hodges (1994). This idea was aired at a meeting in 1984 (Group for Assistance on
Systems relating to Grain After harvest (GASGA), 1984) (R.H. Smith, pers. comm.), and
its development was in part inspired by the discovery that males decrease their pheromone
signal in response to an involatile chemical produced by females but not males (Smith,
1996).
The main critic of this theory so far has been Dr.H.Y.Fadamiro (see Fadamiro,
1995). Fadamiro suggests that the pheromone shut down in response to females is a
general response to overall population density and not necessarily a response specifically to
the presence of a potential mate. Fadamiro’s work has focused mainly on the flight
behaviour of LGB. He consistently found no sex-specific differences in response to
4
aggregation-pheromone signals and concluded that aggregation pheromone was more likely
to function as, ‘a suitable resource location and colonisation signal’, (Fadamiro, 1995); this
may be a fair description of part of the motivation that leads beetles to respond to the signal,
however surely any functional explanation for the existence of a signal needs to address the
question, why do males signal in the first place?
1.3 SEXUAL SELECTION

Sexual selection is defined by Darwin as, “The advantage which certain
individuals have over others of the same sex and species solely in respect of reproduction”,
(Darwin, 1871). It is quite common for one of the sexes of a species to be more
ornamented (either visually, chemically, structurally, or acoustically) than the other and
such ornamentation is often attributed to sexual selection (Cronin, 1991; Andersson, 1994;
chapter 8 this thesis). Classic examples include the elaborate Peacock’s tail and the large
antlers of male deer. Perhaps aggregation pheromone signalling in male LGB owes its
origins and/or current maintenance to similar processes that have produced these other
ornaments?
Two of the earliest mechanisms proposed to result in sexual selection were

female choice (inter-sexual selection) and male contests (intra-sexual selection). More
recently scramble competition (favouring the first males to locate mates) and endurance
rivalry (favouring males who are able to remain reproductively active for an extended
period of time) have been added to the list (Andersson, 1994). The ‘battlefield’ on which
relative male reproductive success through sexual selection is determined has extended to
competition between ejaculates within the female reproductive tract (Parker, 1970).
Historically, females tended to be viewed as passive commodities that males strive to
obtain. Females are now recognised to be far more pro-active determinants of relative male
reproductive success (Eberhard, 1996; Wirtz, 1997). Determining parts of the complex
whole of sexual selection in any one species is a daunting, yet rewarding task. See Cronin
(1991) for a comprehensive review of sexual-selection theory.
1.3.1 Low sexual dimorphism of LGB external morphology

Current morphology can hint at past selection pressures. Externally, adult
Prostephanus truncatus are not strikingly dimorphic in form. The relative difficulty in
sexing this species pays testament to the lack of sex-specific structures, although means of
continuously variable features are different (sex specific body size and size and shape of a
pair of bumps on the head for instance (Shires and McCarthy, 1976)). The horns of male
stag beetles, and the longer antennae and bristles on the legs of males of some
Monochamous species of the Cerambycidae are two examples of features that warrant
more investigation for their possible role in sexual selection (see Hughes, 1981 for study of
a Monochamous species). Females are often larger than males in invertebrates (Andersson,
5
1994). Higher fecundity with increased size of female and better manouvrability used in
mate acquisition with decreasing size for males are cited as two possible explanations for
this difference. Shires and McCarthy’s method for sexing LGB is based on the fact that
females generally have more prominent clypeal tubercules (bumps on the head), which are
spaced further apart in females than males (Shires and McCarthy, 1976). The function of
these bumps is unknown. Perhaps these protrusions are useful during aggressive
encounters. It will be noted in chapter 6 that females are generally more aggressive than
males and appear to use pushing behaviour to choose between mates. Such ‘rejection’
hypothesis for traits that are more elaborate in females than males where no reversed sex
role is inferred are discussed briefly in Andersson (chapter 13, 1994). Boring insects such
as LGB may be particularly constrained against possessing elaborate external features since
they could impede tunnelling.
In this study I will look across the spectrum of possible determinants of male-
fertilization success in LGB to determine the consequences of male aggregation pheromone
signalling. The response of both males and females to different aggregation pheromone
signals is perhaps the first sieve in the behavioural sequence that determines mating
success. The question of whether males can bypass this sieve and cheat has been addressed
by considering the next possible sieves, behaviour on contact (both between the sexes and
between males) and processes within the female tract. Practically nothing was known about
the reproductive biology of this insect, so initially a huge range of determinants of mating
success were possible. This study presents new findings on all levels of the reproductive
biology of this species, from which it is hoped that more accurate hypotheses can be
derived in the future.
1.4 SUMMARY OF CHAPTERS

The idea that aggregation pheromone signals are exploited sex pheromone
signals was first investigated by a review of the available literature on Coleopteran
aggregation pheromones in chapter 2, where the main rival theory, that of a food
conditioning benefit, is also discussed. Chapters 3-7 deal more specifically with
Prostephanus truncatus. Basic patterns of signalling and response are presented in chapter
3, a mandatory prerequisite to a more in-depth study. Chapters 4 and 5 demonstrate that
considerable variation exists between male aggregation pheromone signals both in a
laboratory bioassay and in trapping trials in the field. This variation is perceived by
conspecifics who ‘agree’ which signals are the most attractive, therefore females may mate
non-randomly on the basis of this pheromone signal. The courtship behaviour of
Prostephanus truncatus is described in chapter 6. On contact, females continue to
discriminate between males by aggressively pushing potential suitors away. In common
with the preferences shown to different pheromone signals, different males were ranked in
the same order by different females. The relationship between success in attracting females
6
by signalling and success in physical courtship was then tested. Natural variation in both
traits revealed no consistent relationship between these two possible components of mate
choice. The pheromone signal was then manipulated to show conclusively that this signal is
not a determinant of courtship preferences. Behaviour of adults within an artificial plant
host, recorded using time-lapse photography is also reported in chapter 6. The potential for
sperm competition in LGB is evaluated in chapter 7 along with some assessment of male
ejaculate investment in differing sociosexual environments. Lastly (in chapter 8), the results
are discussed in the context of sexual selection theory and also with reference to the
continuing pest management of LGB.
1.5 GENERAL MATERIALS AND METHODS

1.5.1 Source of insects
All insects used were originally collected from Tanzania in the 1980’s unless
otherwise stated. These insects were provided by the Natural Resources Institute (NRI) and
were maintained on maize for 2-3 years at Reading University and then a further 3 years at
Leicester University, prior to the start of this study in February 1995.
1.5.2 Source of plant material used as insect hosts

Whole maize was supplied by Spillers, Bristol, UK. Maize meal used was
‘Natco medium com meal’ supplied by T. Cholthram and sons Ltd., Middlesex, UK.
Cassava (unfermented) was obtained from Dr. Rick Hodges at NRI and came originally
from Ghana.
1.5.3 Insect cultures

Insects were maintained on whole maize in glass jars (capacity=550ml). All
food sources were sterilized by freezing and stored at -20°C. Each new culture consisted of
approximately 200 adult Prostephanus truncatus placed on half ajar of maize. Jars were
sealed with a plastic screw lid with most of the top replaced by a filter paper held in place
with metal gauze to ensure no insects escaped. All jars were placed on an oil coated surface
to prevent the spread of any insect or mite infestations. No cultures ever became infested
with mites during this study. All cultures were kept in a constant temperature humidity
(CTH) room set to 30 °C and 70% relative humidity. There were occasional problems with
the humidifier and although for approximately 80% of the time conditions were as stated
above, the range recorded in the CTH room was 25-33°C and 40-85% r.h. Cultures will be
affected much less by short-term fluctuations than the sensors in the CTH room. The life
cycle from egg to adult took approximately four weeks under these conditions.
Insects were removed from culture by sieving through brass sieves of mesh
sizes 850jiim and 3.35mm (Endecotts Ltd., London, UK). If the protocol required insects
7
Chapter 1: Introduction
to be reared separately for a time prior to an experiment, then insects were usually reared in
small glass pots with perforated plastic lids (height=5cm, diameter=2.3cm). Unless stated
otherwise, approximately 5cm3of coarse ground maize meal was used to feed these insects.
Demianyk and Sinha (1988) found that a single whole grain of maize was sufficient to feed
one adult for an average life span so the amount of food used in this study was never
limiting.
1.5.4 Experimental conditions

Unless stated, all behavioural observations were carried out in the same CTH
room where the cultures were kept (generally 28-31 °C and 55-75% r.h.). Initially care was
taken to rear some test insects in a separate ‘pheromone free’ incubator, but it became
apparent that this was not necessary for the trials conducted.
1.5.5 Insect handling

Insects were generally moved using a small piece of paper towel. It was found
that if the towel was brushed lightly against the insect’s legs, LGB will cling on and can
easily be moved without the need for forceps that can on occasion damage limbs.
‘Featherlite’ forceps were, however used to transfer insects during weighing as this
required an accurate placement of the beetles on the small weighing platform. During
examination of genitalia required for sexing (see section 1.5.6), a short length of thread
was taped across the tip of the forceps. This thread was pinned across the ventral side of
the upper abdomen of the adult and was used for holding the insect on its back.
1.5.6 Sexing Prostephanus truncatus

Clypeal tubercle size and shape were initially used to sex adults as suggested
by Shires and McCarthy (1976). In this method the sexes are distinguished by looking at
two bumps on the head (clypeal tubercles) that are larger and spaced further apart in females
relative to males. This method is not 100% accurate since variation in this character is
continuous and the female and male range overlaps. It is possible to reduce the rate of error
by only selecting individuals at the extreme ends of the spectrum, but since the extent of
variation between individuals was the subject of some of this work, this selection would be
inappropriate. I decided to sex beetles by looking at a more definitive indicator of their
gender, their genitalia. Females possess a bright yellow, sac-like structure that is revealed
when the last abdominal segment is lifted in response to slight pressure applied to the
abdomen. Likewise, males will extrude their three pronged intromittent organ in response
to this pressure (see section 7.4.1 for description of the intromittent organ). The main
disadvantage of this method compared to that of Shires and McCarthy (1976) is that insects
are more likely to be damaged whilst the pressure is applied. Scholz independently decided
to use genitalia to sex LGB (Scholz, 1997). Scholz specifically investigated the risk of
damage from sexing by everting genitalia and found no significant effect on either mortality
8
or fecundity arising from this procedure (Scholz, 1997). After practice, beetles can easily
be sexed at a rate of approximately 80 per hour if live undamaged adults are needed, or at
about twice this rate if the sex of freshly killed insects is required.
1.5.7 Video equipment

All video recordings were taken using a Vista (Japan) NCL1100 CCTV camera
with a Vista (Japan) CCTV lens, 3.5mm, FI.4. An Hitachi time-lapse video recorder
(model VT-L2000E) was also used.
1.5.8 Statistical analysis

All error bars are standard errors unless otherwise stated. Likewise, all
statistical analysis were performed in SYSTAT 5 for the Macintosh unless otherwise
indicated.
9
Chapter 2: Literature review
CHAPTER 2: LITERATURE REVIEW OF COLEOPTERAN

AGGREGATION PHEROMONES
2.1: DEFINITION
Communication between organisms takes many forms. From a human
perspective watching and listening to others are perhaps the most obvious ways in which
we receive signals. For insects, the subject of this review, the world of smells/tastes has
heightened importance and can relay surprisingly precise information (Butler, 1967).
Karlson and Butenandt proposed the term ‘pheromone’ in 1959 to describe, ‘substances
which are secreted to the outside by an individual and received by a second individual of the
same species in which they release a specific reaction’. This definition serves to distinguish
pheromones from hormones, which are chemicals conveying information within the body
of a single organism, and also from kairomones (attracting exploiters) and allomones
(attracting organisms of benefit), which are chemical messages, intentional or otherwise
between individuals of different species. These definitions are in no way mutually
exclusive, however, and the same chemical may be acting in more than one capacity at any
one time depending on whom you are considering. Semiochemicals is a more general term
encompassing pheromones, kairomones and allomones. Bossert and Wilson classified
pheromones further in 1963 into ‘releaser’ substances and ‘primer’ substances. Releaser
substances are those which induce an immediate and reversible change in the recipient
acting more or less directly on the central nervous system, and primer substances are those
which trigger a more permanent physical change in the recipient, for example retarding
sexual maturity. In this review I will be dealing with releaser pheromones, more
specifically, “substances produced by members of either or both sexes that induce members
of both sexes to aggregate”, termed aggregation pheromones (Borden quoted in Kerkut and
Gilbert, 1985). These are superficially distinct from sex pheromones, which attract just the
opposite sex from that of the signaller.
2.2: OCCURRENCE
Aggregation pheromones as defined by Borden, are fairly widespread in
insects. They are found in cockroaches, social Hymenoptera and many beetles
(Coleoptera). The literature on Coleopteran aggregation pheromones is composed mainly of
various phytophagous insect-pest species (see Burkholder and Ma, 1985). The impact
created by such aggregations has been the inspiration behind much of the funding given
over to this subject. For this reason, it is these insects, the weevils (Curculionidae), the
grain borers (Bostrichidae), the flour beetles (Tenebrionidae), the sap beetles (Nitidulidae),
10
Chapter2: Literature review
the Chrysomelidae, the flat bark beetles (Cucujidae), and the well studied bark beetles
(Scolytidae) that this review concentrates on.
2.3: CHEMICAL NATURE

Pheromones are organic molecules ranging in size from simple benzene
derivatives to longer carbon chains and more complex multiple ring structures. Pheromones
are often a blend of components whose ratio and chirality can be crucial in determining the
response elicited (Vanderwel and Oehlschlager, 1987; Seybold, 1993). Coleopteran
aggregation pheromones are either chemicals sequestered from the host which are often
slightly modified; or are effectively built up from scratch from simpler precursors (said to
be synthesised de novo).
The relatively well studied bark-beetle aggregation pheromones consist mainly

of bicyclic ketals and secondary alcohols. The biosynthetic pathways leading to the
construction of bicyclic ketals are unresolved but the secondary alcohols are thought to be
mostiy derived from host monoterpenes (see Vanderwel and Oehlschlager, 1987 for a
summary). These monoterpenes are often toxic to the beetles that feed on them and it has
been proposed that the biochemical pathways that produce these pheromones may have
initially evolved as part of a detoxification process. The detoxification strategy is generally
geared to converting hydrophobic structures into hydrophilic ones, which are easier to
eliminate. Various isomers of verbenol and verbenone and myrtenol were thought to be
derived from the tree oleoresin component alpha-pinene in bark beetles. More recent
evidence, however, indicates a greater role of de novo synthesis of Scolytid pheromones.
Ivarsson et al., (1993) and Seybold et al, (1995) have both studied biosynthesis of
pheromone in lps species and found that Ipsenol, Ipsdienol and E-myrcenol can be
produced de novo. Myrtenol, another derivative of alpha-pinene, is known to be produced
by lps paraconfusus , Dendroctonusfrontalis, Dendroctonus ponderosae, and Dryocoetes
confusus. Although myrtenol has so far no known pheromonal function in lps species, it is
a multifunctional pheromone of Dendroctonus frontalis and an aggregation pheromone of
Dryocoetes confusus (Discussed in Prestwich and Blomquist, 1987).
Bostrichid aggregation pheromones are relatively simple compounds containing

9-12 carbons possibly constructed de novo and not obviously sequestered from the host
(Prof. D. Hall pers comm.). So far the aggregation pheromones of Cucujid grain beetles
have been found to be macrolides of either terpenoid or fatty acid origin or possibly by de
novo synthesis (see Vanderwel and Oehlschlager, 1987; and Vanderwel etal., 1992). Both
Tribolium castaneum and T. confusum (Tenebrionidae) use 12 carbon alkanals as
aggregation pheromones. Rhynchophorus species of weevil have been found to use
various alcohols as pheromones (Jaffe et al., 1993; Weissling et al., 1994). Maize and rice
weevils (Sitophilus zeamais and S. oryzae ) both use the same compound which has the
11
typical 1,3-oxygenation pattern characteristic of polyketides (see Vanderwel and

Oehlschlager, 1987). Boll weevils were shown by Mitlin and Hedin (1974) (quoted in
Prestwich and Blomquist, 1987) to incorporate radio-labelled acetate, mevalonate and D-
glucose into all four components of grandlure thus showing that this pheromone could be
constructed de novo. Similarly, Cryptolestes ferrugineus (Rust red grain beetle) was
shown to use both acetate and mevalonate for the construction of a terpenoid pheromone
(quoted in Prestwich and Blomquist, 1987). However, both these species can also derive
these pheromones more directly from dietary components.
Micro-organisms residing in the guts of insects have sometimes been shown to

convert dietary chemicals into semiochemicals. For example, a Bacillus cereus strain has
been isolated from lps paraconfusus. Experiments using externally-sterilised eggs that were
then grown in axenic conditions, and adults treated with antibiotics have indicated that both
insect and microbes are involved in ipsdienol production by male lps paraconfusus. No
such microbial influence was shown for Dendroctonus ponderosae (work quoted in
Prestwich and Blomquist, 1987). The micro-organisms do not necessarily always add to
pheromone output, indeed axenically-reared male lps paraconfusus and female
Dendroctonus ponderosae contain significantly more alpha-pinene metabolites than wild
beetles. Increased pheromone production in microbe-suppressed insects might simply be
due to healthier insects or competition for pheromone substrates between microbe and
insect. It has also been recognised that microbes may convert attractive compounds made by
the insect into anti-aggregation compounds, which could play a significant part in
terminating the attack on a host.
To summarise, so far pheromone origins have been proposed from fatty acids,
polyketides, and terpenoids. Many unknowns still exist particularly for low molecular-
weight pheromones. Much of the work performed so far has been on bark beetles, but
results of studies in a few other beetle groups can be found.
2.4: BIOLOGICAL FEATURES

In this section, biological attributes of aggregation pheromones are given as
generalisations (in italics) followed by a justification and expansion.
Pheromone produced only when signaller is on the host:

Aggregations of insects often form away from an appropriate food source,
usually for the purpose of mating. Alternative localities include potential adult emergence
sites, potential oviposition sites and conspicuous features of the landscape like hill tops
(Thornhill and Alcock, 1983). All the aggregation pheromones reviewed here are usually
produced in the presence of a suitable food source. All the food sources reviewed also
double as sites of oviposition, perhaps their more important role since often more of the
host is consumed by the developing larvae than by the mature adults.
12
The necessity of a host as a stimulus for pheromone production has been

specifically demonstrated in Chrysomelidae (Peng and Weiss, 1992), Nitidulidae (Bartelt
and James, 1994), Bostrichidae (Mayhew and Phillips, 1994), and Scolytidae (Vite and
Pitman, 1968). With the exception of Peng and Weiss (1992) these studies all assess
pheromone production directly by collection of volatiles and not via the response of
conspecifics. This response may be influenced by the presence of food directly as well as
its effect on pheromone production (see next sub-heading). Peng and Weiss (1992)
controlled for this effect by separating potential producers from the food source with a
screen. Faustini et al., (1982) and Rochat et al, (1991a) have obtained similar results in
their studies in Curculionidae but without such a control being included.
Even the distinction insinuated between host odours and insect-produced

pheromones is far from obvious since many pheromones are derived more or less directly
from host chemicals ingested by a feeding insect (see previous discussion). Pheromones
may be assembled from many basic building blocks by the signaller. Where the pheromone
is derived more directly from the host-plant chemistry, presence of the host is obviously
required for pheromone production. Interestingly, the host plant also seems to be required
for production of aggregation pheromone even when the pheromone is truly being
synthesised de novo by the insect. It is easy to imagine that some pheromone systems
could have evolved from signallers enhancing cues given out by host plants for their own
gains, perhaps initially by simple physical damage, and later by synthesis of these
chemicals by the insect itself. Such, ‘sensory trap’, ideas are discussed by Christy (1995).
Christy notes that in Lepidoptera, ‘unlike female odours, male odours usually are similar to
or the same as the odours of adult or larval food plants’. Thus, for these species at least,
sensory traps are a tactic more often employed by males during mate attraction than females.
There is evidence that some Dendroctonus species of Scolytids do not require

feeding as a cue for aggregation-pheromone production (see Wood, 1982). In these cases
pheromone release from females can be stimulated by a sonic signal from the male.
Response to pheromone is often increased by confirmation o f the presence o f food via

perception o f food volatiles:
Potential responders to a pheromone signal may assess the existence of a food
source before responding, by using host volatiles (for a review see Visser, 1986). The
importance of the pheromone signal to be in the context of such volatiles varies from very
strong synergism such as that found in Nitidulid beetles (Bartelt et al., 1993a), to as yet no
detectable influence (Pierce et al., 1995 for Conophthorus species of Scolytidae, but see
enhancing effect noted by Birgersson et al, 1995). Bartelt et al., (1993a) found that
pheromone alone attracts only 1-29% as many beetles as the combinations with dough
(source of food volatile), and dough alone only 0-2.9% as many as in combination. Here it
is volatiles produced by the action of yeast on the host rather than volatiles emitted by a host
13
alone that seem to be most influential. Other synergistic influences of pheromone/food

volatile combinations have been described in the Curculionidae (Walgenbach and
Burkholder, 1986; Weissling etal., 1994; Giblindavis et al, 1994; Gries et al, 1994b;
Rochat et al, 1993; and Phillips et al, 1993) and Scolytidae (Pitman, 1969; Byers et al,
1990). Other patterns of influence include additive effects (Renwick and Vite, 1969) and
density-dependent but ill-defined effects (Dowdy et al, 1993), and Jaffe et al, (1993)
proposes an interaction that is distance-dependent. Petroski and Vaz (1995) proposed that
the pheromone and co-attractant bind in close proximity on the receptor surface. This
evidence comes from molecular modelling of 26 host-type volatiles with pheromone from
Nitidulidae. A recent review of the influence of host chemistry on the biology of attracting
pheromones is given by Landolt and Phillips (1997).
Host volatiles are used as a tool in pest management. The use of host volatiles
in combination with pheromone components in traps is reviewed in Faustini et al, (1990).
Host volatiles could also be used to confuse beetles. Inappropriate host volatiles have been
shown to interrupt responses to true host volatiles and to an aggregation signal in bark
beetles (Dickens et al, 1992).
Produced by species with a relatively long lived adult stage:

According to Burkholder and Ma (1985), “two general types of communication
and reproductive strategies characterise stored product insects”. They propose that insects
are either short lived, do not feed as adults, and produce sex pheromones; or they are long
lived, have to feed as adults, and produce aggregation pheromones.
The food source is capable o f supporting many individuals:

Here the data available are, unfortunately biased towards studies on insect-pest
species. Therefore food sources are often artificially large, a good example being a grain
store. I have found no studies on beetle species that generally feed on very small patches of
food and produce aggregation pheromones. Such situations, however, could not result in
the large groups of insects which tend to stimulate research funding, so the possibility of
this occurring cannot be dismissed.
The identity o f the signaller is sex specific:

Both males and females have been found to produce aggregation pheromones.
Male-produced aggregation pheromones have been described in Bostrichidae (Dendy et al,
1991; Khorramshahi and Burkholder, 1981); Cucujidae (White e ta l, 1989; references in
Phillips, 1994); Curculionidae (Budenberg et al, 1993; Faustini et al, 1982; Hibbard and
Webster, 1993; Jaffe et al, 1993; Nielsen and Jensen, 1993; Patrock et al, 1992; Perez et
al, 1994; Rochat et al, 1991b and 1993; Walgenbach et al, 1983; Weissling et al, 1994);
Nitidulidae (Bartelt et al, 1993b; Petroski et al, 1994); Tenebripnidae (Obeng-Ofori and
Coaker, 1990); Scarabae (Gries et al, 1994a), and some genus of Scolytidae including lps
14
(Akers et al, 1993; Borden, 1967; Miller et al, 1991; Seybold et al, 1995; Byers et al,
1990; Teale et al, 1991a; Wood, 1982; Zuber et al, 1993); Polygraphus (Bowers and
Borden, 1990) Dryocetes (Camacho e ta l, 1994), andPitogenes (Byers, 1993).
Females have been described as the initial signallers in several genera of

Scolytidae including Dendroctonus spp. (Agosta, 1992; Renwick and Vite, 1969; Wood,
1982), Scolytus spp. (Wood, 1982) and also Trypodendron lineatum (Raffa et al., 1993).
Usually female weevils produce sex pheromones but an aggregation pheromone has been
described from female cabbage seed weevils, Ceutorhynchus assimilis (Evans and
Bergeron, 1994).
Mostly only one sex produces aggregation pheromones. Where both sexes
signal it is mostly with different chemicals. Cases where both sexes signal with the same
chemicals are rare and in the case of the Cucujid beetle, Ahasverus advena the rate of
production differs with males producing at least four times as much as females (Pierce et
al, 1991).
The pattern o f response is sex specific:

Aggregation pheromones by definition attract both males and females. There
are, however, many sex-specific differences in response to aggregation pheromones
(review of Scolytidae, Raffa et al, 1993). Sex specific differences can be complex to detail
since the behaviour these pheromones elicit appears to be very variable both between
individuals and for any one individual at different times. Thus any comparison between the
sexes in terms of their response can give widely varying results very much dependent on
the design of the experiment. Experiments that record the number and categories of beetles
attracted to a pheromone source in the field can give insight into which conspecifics a
signaller may expect to attract. Attraction depends on many factors including differential
sensitivity of the responders to pheromone and which portion of the population is
dispersing.
Zuber and Benz (1992) found a higher proportion of males in traps at the
beginning of a flight period in lps typographus thought to arise from sex specific patterns
of dispersal. A similar situation was found by Chapman (1966) for Trypodendron lineatum
, and by Rudinsky (1963) for Dendroctonus pseudosuage where additionally female
numbers peaked late in the season in synchrony with re-emergence for “second brood”
formation. Polygraphus rufipennis showed no sex-specific catch differences in the Spring
and Summer, however more of the opposite sex to the signaller were caught in Winter
(Bowers and Borden, 1990). Teale et al, (1991a) found a seasonal variation in response to
ipsdienol by lps pini. Ips typographus and lps pini response to pheromone was also found
to be affected by nutritional state (Nemec et al, 1993 and Gast et al, 1993 respectively).
Mating status can also affect response (see next sub-heading). Laboratory-based studies can
15
be designed to look specifically for any patterns of differing sensitivity to pheromone,

which may give us further insight into the motivation behind the behaviour of responders.
Some studies have found no difference in response between the sexes (Field
studies: Patrock et al., 1992; Peng and Weiss, 1992. Laboratory studies: Dowdy et al.,
1993; Obeng-Ofori and Coaker, 1990; and Walgenbach et a l 1983). The opposite sex to
the signaller usually shows the greater response where there is a difference in response
between the sexes: Field data: Weissling et al., (1994); Krausseopatz et al., (1995); Evans
and Bergeron (1994); Renwick and Vite (1969). Laboratory data: Borden (1967); Weissling
et al., (1994); Evans and Bergeron (1994); Faustini et al., (1982); Gast et al., (1993,
Bowers and Borden 1990. The results obtained by Obeng-Ofori and Coaker (1990) for
Triboliwn castaneum are the expection where male pheromone attracted more males than
females in a laboratory bioassay.
Schlyter et al., (1987) suggested that males responding to male produced

pheromone may avoid high concentrations of the signal and land a short distance from the
source to avoid inter-male competition. Some sex-specific patterns of response recorded
from trap-catch data may therefore arise from sex specificity in landing behaviour rather
than from differences in the numbers of insects orientating towards the signal.
Sex-specific differences have also been found at the level of pheromone

perception. Chambers et al., (1990) found females of the Cucujid beetles, Cryptolestes
ferrugineus and C. pusillus had larger electroantennogram (EAG) responses than males
when exposed to the male-produced pheromone components. Faucheux (1994) has even
found structural differences between the sexes with respect to their sensilla cells. EAG
results are just one level of influence on behaviour, and studies by White and Chambers
(1989) show marked sex-specific behavioural differences where no differences in EAG
patterns were found.
Discrepancies between the response of the sexes can be further complicated

when different components of pheromone blends skew the sex ratio in different directions.
As we find out more about pheromone systems it is likely that more components will be
revealed and an ever more complex pattern of response will become apparent. Ips pini and
Ips typographic are good examples where complicated patterns of response are specific to
the sex of the responders (Gast et al., 1993; Schlyter et al., 1987; Teale 1991). In another
Scolytid, Dendroctonus brevicomis, at least one attractive pheromone component is
produced by each sex that preferentially attracts the opposite sex; the combined blend results
in an approximately equal sex ratio (Agosta, 1992). The Saw toothed grain borers,
Oryzaephilus surinamensis and Oryzaephilus mercator have a multi-component pheromone
whose blend also determines the degree of sexual dimorphism of response (White and
Chambers, 1989 and refs, therein). The Tenebrionid, Triboliwn castaneum shows sex
16
specific differences of response than can be altered by changing the ratio of the components
(Rangaswamy and Sasikala, 1991). The enantiomeric ratio of single components can also
have a sex-specific effect on response (Salom et al, 1992). The ratio of pheromone to host
volatiles may also influence the sex ratio of beetles attracted. In Dendroctonus ponderosae
low trans-verbenol:resin ratio attracts mainly females where a high ratio results in a male
bias (Quoted in Raffa et al., 1993).
Production and/or response to pheromone often decreases with mating experience:

Arrival of the opposite sex and/or mating experience have often been found to
decrease pheromone production. This is not always the case, for example mating was not
found to affect pheromone output by males of the maize weevil, Sitophilus zeamais
(Walgenbach et al., 1983) although boll-weevil males decrease aggregation-pheromone
production in response to a single mating with females (Dickens and Wigul, 1987). As yet,
no effect has been demonstrated in the Bostrichid, Rhyzopertha dominica, (Dowdy et al.,
1993; Mayhew and Phillips, 1994), however in a related species, Prostephanus truncatus,
pheromone production by the male can be decreased dramatically by exposing the signaller
to females, even if indirectly in the form of grain previously infested by females (Smith et
al., 1996). Both Polygraphus rufipennis and Ips confusus males showed decreased
pheromone production in the presence of females (Bowers et al., 1991; and Borden, 1967
respectively). Production decreases in direct correlation with the number of females per
male in Ips confusus (Borden, 1967). Decline in female attractivity on addition of males
has been demonstrated in Chapman’s studies on Trypodendron lineatum and Rudinsky’s
study of Dendroctonus pseudotsuage (Chapman, 1966; Rudinsky, 1963).
Some sex-pheromone systems also exhibit pheromone shut down in response

to mating of the signaller. Pheromone titre produced by the female dropped dramatically
within just two hours of mating in the moth, Heliothis virescens (Raina, 1990), and was
associated with the female becoming unreceptive to further matings. In H. virescens
receptivity and pheromone signalling can resume after a one-day refractory period, although
in some moths females only mate once and females never resume signalling (discussed in
Raina and Stadelbacher, 1990).
Mating status can also be correlated to the level of response to pheromone.

Quoting from Borden (1967) when discussing Ips confusus, “reproducing females show a
greatly reduced response to male attractant”. Walgenbach et al., (1983) found that only the
female’s response to pheromone decreased with mating and the male’s response remained
high whatever the mating status. No differences have been found between the responses of
virgins versus non-virgins in Bostrichidae (Obeng-Ofori and Coaker, 1990; Dowdy et al,
1993), or Tenebrionidae (Obeng-Ofori and Coaker, 1990), and Rochat et al., (1991a)
found no obvious effect for the American palm weevil, Rhynchophorus palmarum. Gast et
al., (1993) noted a general pattern where gonad development in adults was significantly
17
correlated to response to pheromone and so, “proportionately more responders were

sexually mature than in the total population”.
2.5: ANTI-AGGREGATION PHEROMONES

The growth of aggregations is often limited at least to some degree by repellent
pheromone signals and/or the cessation of signalling. Repellents can be distinct chemicals
from the original pheromone or different blends, or concentrations of chemicals that in
another format, are attractive (see discussion in Schlyter et al., 1987). For example the
blend of endo- and exo-brevicommin determines whether it is an attractant or repellent to
responders of Dryocetes confusus (Camacho et al, 1994).
The trigger for anti-aggregants is thought to be a function of population density

(Agosta, 1992). It is also often found that the rate of production of attractive pheromone per
beetle decreases with the number of signallers in an aggregation. For instance, Australian
sap beetle, Carpophilus davidsoni males in groups of 50-60 emitted a peak level of 0.09jLLg
per beetle per day compared with a single male emission rate of over 3fig per beetle per day
(Bartelt and James, 1994). Amount of pheromone per male decreased when population
density was artificially manipulated in the Bostrichid Rhyzopertha dominica in an
experiment using mixed cultures (Dowdy et al, 1993; and Mayhew and Phillips, 1994).
Pierce et a l, (1991) found that pheromone production of the Cucujidae, Ahasverus advena
was, ‘barely detectable’, in the highest population densities of mixed sex cultures tested.
Unfortunately it would be very difficult, if not impossible to monitor the breakdown of
pheromone production across many co-habiting individuals. It is not known whether such
decreases in pheromone production arise from some individuals completely ceasing
production or a more general lowering of production by all individuals. In common with
attractants, response to repellents can vary with their release rate (Bertram and Paine, 1994;
Miller et al, 1992), the season (Devlin and Borden, 1994), and many other influences.
Unlike the attracting pheromones, aggregation inhibitors are usually produced

by both sexes. For example both males and females of Dendroctonus brevicomis (Bertram
and Paine, 1994) and D. frontalis (Payne, 1992) emit verbenone, a common aggregation
inhibitor in Scolytidae. 3,2-MCH has inhibitory effects for D. pseudotsuage and is
produced by both sexes (Wood, 1982; Ross, 1994). Both males and females of Ips
paraconfusus produce cis-verbenol (Wood, 1982), which acts as an inhibitor of male
produced aggregation pheromone components (Akers et a l, 1993). Anti-aggregation
pheromones are generally much less species specific than their corresponding attractants.
As already mentioned verbenone is common amongst the Scolytidae, as is Ipsdienol
(references above and; Safranyik, 1992; Shore etal., 1992; Paine and Hamilton, 1991;
Devlin, 1994; Miller and Borden, 1992; and Byers, 1993). It is proposed that the higher
generality of these compounds between species allows them to function to limit both inter as
18
well as intra-specific competition. Indeed the pheromone of one species often has an
inhibitory effect on other sympatric species (Schlyter and Anderbrandt, 1993; Miller and
Borden, 1992; Byers, 1993).
Anti-aggregation compounds have obvious potential to be applied to pest

management. They can be used to disrupt colonisation (Payne et al, 1992) and they are
often used in conjunction with attractants in a ‘push-pull’ strategy that can concentrate the
pest away from high value commodities (Ross and Daterman, 1994; Smart et al, 1994).
2.6: NON ADAPTIVE EXPLANATIONS

The problem to be discussed is, ‘why do insects send these signals?’. Both
adaptive and non-adaptive explanations might explain signallers’ behaviour. We might
expect the behaviour of responders to be adaptive, if not at present then surely at some
point in the past. I suggest this because orientating towards pheromone sources may be
expected to be costly in terms of both time and energy, particularly if flight is involved. In
contrast, signalling could be of little cost, unavoidable or in fact costly to prevent. Many
pheromones, as already mentioned are derived from host compounds and may be waste
products of metabolism, in particular digestion. To eliminate the emission of such
compounds following feeding may represent a significant cost to the signaller (Renwick and
Hughes, 1975 quoted in Alcock, 1982). Here the signaller may actually suffer from
advertising its position, and so no benefit from attracting conspecifics is necessarily
inferred. The finding that generally only one sex signals implies that pheromone production
is at least avoidable in these cases. If the pheromone is synthesised de novo as is indicated
to be the case in some beetles studied (Seybold et al, 1995; Ivarsson et al, 1993), then
production is likely to represent an energetic cost of some description, and indeed Schlyter
and Birgersson (quoted in Schlyter and Birgersson, 1989) estimates that male Ips
typographus expends the equivalent of 2% of its body weight per day when emitting
pheromone.
2.7: ADAPTIVE EXPLANATIONS

It is tempting to look on the scale of the group when looking for adaptive
explanations for the formation of aggregations, after all, it is a phenomenon involving many
individuals (see Shorey, 1973). Explanations like, "...because the food resource may be
either ephemeral or not available until the tree dies, these beetles have evolved a pheromone
that elicits behaviour resulting in aggregation of the population on the new host.." are put
forward (Wood, 1982). Here I stress that potential explanations should be described in
terms of individuals to avoid wrongly categorising such explanations as group selectionist.
Therefore I would start rephrasing Wood's proposal to read, "...a beetle that locates a host
might benefit from advertising its location because...", followed by, "...beetles may move
towards/away from this signal because...". Notice how the signaller cannot necessarily
19
dictate the response of the receivers of the signal and consequently may expose themselves
to exploitation. Alcock (1982) has led the way in demanding explanations phrased in terms
of individuals for the existence of bark beetle pheromone systems. Here I would like to
extend this method to encompass other Coleoptera.
2.8: COSTS OF SIGNALLING

Costs of signalling are not limited to energetic costs since the signal may also
create costs from its influence on other species. Potential competitors and/or predators can
be attracted. Predators from a wide range of taxa have been shown to follow aggregation
signals (Prestwich and Blomquist, 1987; Wood et al, 1968; and refs, in Wood, 1982) and
can be a selection pressure capable of changing pheromone structure/ composition. Ips pini
varies the enantiomeric blend of ipsdienol from place to place and from year to year, in
concordance with changes in response of its predator Thanasimus dubius, this possibly
represents predator/prey co-evolution in action (Herms et al, 1991; Raffa and Dahlsten,
1995). The possibility of such a system leading to polymorphism in respect to pheromone-
production characteristics has been modelled for the European com borer moth
(Guldbrandtsen, 1995). Interestingly in this case just one-locus seems to determine
pheromone blend in females, which can be traced using simple mendelian genetics.
Increased intraspecific competition for host resources is a potential cost

incurred by signallers. As discussed above, however, hosts of these species are typically
large and able to support many individuals. Where attracted conspecifics are the same sex as
the signaller, a cost of increased competition for mates may be incurred. Mate availability is
usually most limiting for males for whom this cost may be most significant.
2.9: BENEFITS OF SIGNALLING

The outcome of aggregation pheromone signalling systems is the formation of a
cluster of individuals of both sexes on a food source. Potential benefits received by the
individuals who invest in signalling could be better defending capacity against predators,
overcoming the resistance of a potential host, and increased opportunity to mate and leave
offspring (Thornhill, 1983).
Defence against predators has been proposed as an adaptive advantage of

aggregating in the aposematically coloured beetle, Lycus loripes (Eisner and Kafatos,
1962, quoted in Thornhill, 1983).
Collective overcoming of host defences is a convincing function of aggregation

in some species of bark beetles, particularly those that attack healthy trees (Wood, 1982).
When a pioneering beetle has located a potential host it is proposed that it will either not be
able to feed and/or oviposit efficiently until the host is transformed or ‘conditioned’ by the
combined action of many beetles. Such a conditioning benefit is often inferred, but rarely
20
Chapter 2: Literature review1
has it been demonstrated directly (with the exception of Raffa and Berryman’s study of
Dendroctonus ponderosae quoted in Alcock, 1982). An early study of Dendroctonus
ponderosae demonstrated a possible food conditioning motivation for signalling, since a
higher rate of pheromone production was observed when females were introduced to logs
with higher oleoresin content in exudent (a method of host defence) (Vite and Pitman,
1968).
Conditioning may involve the inoculation of the host with fungal pathogens as
is the case in some Scolytidae including Dendroctonus pseudotsuage. Scolytus multistriatus
is also thought to inoculate the tree with pathogens e.g. Dutch Elm disease (Wood, 1982).
Conditioning can simply involve the mechanical action of many beetles. ‘Mass attack’, the
situation where an aggregation forms very rapidly, could be a method of killing a tree.
Alternatively, signalling may be an incidental product of feeding, and the ‘mass attack’
often observed may simply arise as individuals race to take advantage of a comparatively
short window where the host is relatively safe in terms of lowered defences and yet still not
totally used up by those beetles already present (see Alcock, 1982). Although some
Scolytid beetles are capable of attacking healthy trees, these are usually only colonised
when stressed or dying trees are unavailable, and the majority of species will only colonise
weakened or even dead hosts. Dendroctonus ponderosa and D. brevicomis can attack
healthy trees but beetles have been found to be preferentially attracted to stressed trees
(Wood, 1982). The Palmetto weevil, Rhynchophorus cruentatus is specifically attracted to
volatiles from dying palms, its host (Giblindavis et al, 1994).
In the stored product environment where the host resource is commonly made
up of much smaller units of food, grain for instance, conditioning could take the form of
changed humidity and substrate consistency from the action of many beetles. Beetles could
potentially share the cost of burrowing through hard seed coats and use common entrances
into a host. However, the benefit of this must be limited by the size of many grains
inhabited by these beetles. A single ovipositing female may utilise more than one, and not
just part of one grain. For example in the grain borer, Prostephanus truncatus a female will
distribute her eggs on average in 12.5 grains (SE = 4.17, n=30) when ovipositing alone
(Li, 1988). However, Fadamiro (1995) still proposes that the individual cost of boring can
be decreased by feeding with more conspecifics. Walgenbach et al, (1983) proposed that
chewing through the seed coat of a grain kernel required much energy in Sitophilus zeamais
aggregations since females require some 30 minutes or more to make a hole in a wheat
kernel large enough to deposit an egg. They observed that when a weevil was feeding on a
fresh kernel of wheat one or more companions were often present and all the grain was
consumed before a new grain was burrowed into. I have yet to find a study that looks
specifically at the division of cost of burrowing between individuals and obviously if a
21
beetle does not defend the food resource it has created access to there is potential here for
exploitation by conspecifics.
Another potential benefit of high adult numbers could be increased offspring

survival from a general increase in the amount of food found as dust/flour outside the
grains. The survival of juveniles may be largely determined by the environment which they
find outside the grain. Offspring of species who oviposit within grains like Prostephanus
truncatus may show particularly high larval mortality if they happen to leave the grain
whilst tunnelling for food. For this species risk of death may be expected to be highest
when the adults are ovipositing on smaller grains since the larvae are more likely to fall out
(suggested by Li, 1988). Li has found no evidence for such a collective-conditioning
benefit in this context for the Larger Grain Borer. For initial adult densities ranging from 6-
26 pairs per 60g of maize the average number of offspring per female emerging was never
found to be higher for an increase in adult density (Li, 1988).
Aggregation pheromones may be used as sex attractants and the benefit

incurred to signallers may not be one associated with a general increase in population size,
more an increase in the availability of mates. At least some studies assume that the main
benefit of aggregation pheromones is increased mating opportunities for the signaller
(Lewis and Austad, 1994; Landolt and Phillips, 1997). Sex pheromones, i.e. those that
attract only the opposite sex have been well studied (see a review for moths by Svensson,
1996). Initially signallers may benefit from attracting members of the same sex through
increase in the efficiency of signalling, but these beetles may be more accurately described
as a cost of signalling since they are also potential competitors for mates. This is thought to
be the situation found in some field crickets described in Alcock (1982) who signal by
chirping.
Where males are the signallers it is feasible to imagine that females could
choose between mates on the basis of these signals if they are variable, and therefore some
pattern of assortative mating could be governed by these pheromones. Where different
pheromones are used for short and long range signalling it may be predicted that the shorter
range components will be most influenced by sexual selection since it is at this point that a
female is most likely to encounter more than one signal and has most opportunity to make
comparisons. It has now been shown that female preference in Ips pini is actually
correlated to the blend of enantiomers of ipsdienol produced by individual males which
leads to this feature being a main determinant of the pattern of assortative mating in this
species (Teale et al, 1994). Alternatively the signaller may be exercising a form of mate
choice by selecting only those insects that have survived the dispersal phase and have
successfully orientated to the pheromone signal (discussed in Raffa et al., 1993).
22
2.10: COSTS AND BENEFITS OF RESPONDING TO

PHEROMONE
Beetles may respond to a pheromone signal as it indicates the location of a host
usually used for both feeding and oviposition, and potential mates. Host distribution and
defences as well as any pattern of mate choice determined by pheromone characteristics may
be predicted to influence the propensity of beetles to follow pheromone plumes instead of
locating an unused host. Also crucial to this choice may be the nutritional and reproductive
state of the insect. A dispersing male low on food reserves that feeds on a rare host would
surely gain most from orientating to pheromone in contrast to a well fed, impregnated
female who utilises a very common host.
2.11: HYPOTHESES DISCUSSED IN TERMS OF THE

BIOLOGICAL FEATURES OF AGGREGATION
PHEROMONES
In this section, I return to the generalisations made in section 2.4 and consider
the specific questions that arise in relation to adaptive functions of aggregation pheromones.
Pheromone produced only when signaller is on the host:

Collective conditioning of a host is a benefit which will obviously only be
realised if the signaller is close to such a host. Mate attraction need not necessarily occur at a
host (Thornhill and Alcock, 1983). However, as discussed above, the chemical nature of
the pheromone may constrain would be signallers to signal only when they have recently
been feeding, and/or the importance of the pheromone signal to be in the context of host
volatiles may constrain signallers to signal only when on a host. In the species considered
here the host is a food source for both adults and developing offspring. Any delay between
mating and oviposition while a female locates such a host will potentially increase the risk
of females mating again before fertilizing her eggs and so promoting sperm competition and
possible loss of paternity for the signaller. See chapter 3 (this thesis) for an assessment of
whether LGB males continue to signal when removed from food, and chapter 7 for an
assessment of the potential for sperm competition in LGB).
Species are long lived as an adult:

A common character of all these species is that they are comparatively long
lived as adults, and colonise hosts which once found/conditioned, can potentially support
many individuals (noted by Burkholder and Ma, 1985). I believe that it is these
characteristics which have facilitated the evolution of aggregation pheromones and
ultimately explain why, if their function is primarily sexual, con-specifics of the same sex
as well as potential mates are attracted.
The food source is capable of supporting many individuals:
23
Fairly long term aggregations (longer than a generation time) would only be
observed if the host can support such a group. The cost of attracting conspecifics in terms
of increased competition for resources will hypothetically become more and more
significant if host sizes were reduced assuming, perhaps wrongly, that insects would not
leave the aggregation to forage. Collective host conditioning would surely only feature
when all insects investing in overcoming host defences are able to gain from subsequent
utilisation of that host. The requirement of a large host is less obvious if these signals serve
to attract mates. However, if for whatever quirk of history they are used primarily as a
method for beetles to increase their rate of copulation, then this kind of system would surely
be the perfect setting for promotion of sexual selection. The opportunity for male-male
competition and female choice being greatest when many individuals are grouped together
in this way.
The identity o f the signaller is sex specific:

Sex specificity of signalling patterns tells us that there are or have been some
sex specific selection pressures which have given rise to this system. The most obvious
differences between the sexes are those directly concerned with reproduction. However,
these differences can have far reaching effects on the rest of the biology of an organism.
The signalling sex sometimes shows a greater response to host volatiles (Gast et al., 1993)
and thus sex-specific signalling could reflect sex differences in dispersal. So although we
cannot assume that signalling is directly involved in reproductive behaviour, perhaps this is
where we should look first for an explanation. Males are the signallers in some species and
females signal in others, what are the differences between these species which could
account for this? Males have the potential to gain most in reproductive output from
increased numbers of mates particularly if they invest little each offspring i.e. provide no
parental care and low amounts of nutrition in the ejaculate. Females could benefit by
promoting all forms of sexual selection by initiating an aggregation, but will be far more
limited in the number of offspring she can produce by energetic considerations.
Perhaps the key characteristic in the less common situation where females are
the signallers is that these are the species that are most likely to be benefiting from a host
conditioning effect. Specifically the Dendroctonus and Scolytus species of Scolytidae,
many of whom are known to kill healthy trees and require this for successful egg laying
(Wood, 1982). However, in the Scolytid species, Trypodendron lineatum, the female
produces the aggregation pheromone, yet this species is not considered to be particularly
aggressive and is not known to actively kill trees.
The pattern o f response is sex specific:

The finding that the opposite sex to that of the signaller often shows a greater
response to pheromone leads to these aggregation pheromones having more in common
with conventional sex pheromones. See chapter 4 and 5 (this thesis) for an assessment of
24
sex specificity of response in LGB: in a laboratory bioassay (chapter 4); and in the field
(chapter 5).
Production and/or response to pheromone often decreases with mating experience:

This finding is perhaps the most convincing evidence that these pheromone
systems are intimately tied to sexual behaviour. Alternative explanations are possible,
Fadamiro (1995) has put forward the idea that male signallers use the presence of females,
detected via an involatile chemical, as an indication of population density, and thus the
observed decrease in pheromone production on exposure to this chemical need not be
explained in terms of any mating function.
Benefit from signalling may undulate quite rapidly for any single signaller. As
mates arrive pheromone production may be cut off for a while until the signaller is free to
remate again. The number of matings optimal for any individual will obviously vary both
between the sexes and between species. Pheromone production should really only be
sustained for polygamous males with low investment per female. Polygamous males of the
maize weevil, Sitophilus zeamais do not show decreased pheromone production with
mating (Walgenbach et al., 1983). High investing males, like Ips confusus, who construct
the nuptial chamber and actually guard this against intruders of both sexes, might be
expected to decrease pheromone production as he acquires the optimum number of mates he
can support. This was found to be the case (Borden, 1967). Decrease in female production
of aggregation pheromone after the arrival of males has not been specifically demonstrated
since the decline in attractivity recorded could be due to male produced repellents as
suggested by Nijolt (1973) for Trypodendron lineatum.
Anti-aggregants:
The point where production of aggregation pheromones switches to the
production of anti-aggregants could be used to infer exactly when and for which individuals
the costs of encouraging conspecific aggregation outweigh the benefits. All the possible
benefits discussed here derived from the formation of an aggregation will tend to be counter
balanced by the costs of sharing resources, particularly as the aggregation grows and the
host patch is depleted. Since these characteristics are common to all hypotheses it might be
unrealistic to use them to distinguish between possible benefits gained. Evidence of
increased intraspecific competition at higher beetle densities has been found in
Dendroctonus ponderosa where there is a strong negative correlation between density of
attacks in cut logs (i.e. defenceless ones), and pupal production per female (Raffa and
Berryman (in prep.) quoted in Alcock 1982). Raffa et al., (1993), suggests that, “inhibitory
pheromones may function as ‘pre-rivalry’ signals in male-male interactions, in that both
signaller and responder benefit from avoiding rivalry”, (see also Pierce et al., 1995 and
Miller and Borden, 1992). The honesty of such signals may be ensured if response to them
is dependent on their concentration and this is a truly constrained by the number of beetles
25
signalling. Dose-dependent inhibition of attractant by verbenone has been found in three

Scolytid beetles where higher doses give most inhibition (Miller et al, 1995).
2.12: FUTURE DIRECTIONS

I have discussed some of the possible selection pressures that may have shaped
aggregation-pheromone systems, the two most likely being host conditioning and mate
attraction. We may never know what happened in the past, but perhaps we can determine
which of these influences are acting to maintain these systems today. Below I outline areas
of further study that could contribute to this discussion.
One potentially profitable approach would be to apply the comparative method

to this topic. A study of patterns of signalling correlated to the life history strategy and
ecology of different species may yield interesting results. I have already mentioned a
possible starting hypothesis, that species in which females are the signallers may be those
colonising hosts with a relatively high capacity to resist attack. Similarly, a review of the
mating systems of species that have aggregation pheromones might highlight interesting
trends. The Scolytidae would be a good start for such a study. Kirkendall (1983) has
reviewed the mating systems of this incredibly diverse family. The data available are
limited, however, they do suggest some associations. Species where males produce
pheromones that have been shown to attract both sexes are, without exception in the sample
reviewed, the initial colonisers and they are harem-polygynous. This means that males in
gallery systems are usually found with more than one female. The shut down of pheromone
in response to arrival of mates appears to be variable between species of the same family
and the distribution of this character among species could also provide further insights.
A contrasting approach would be to study any variation between individual

signallers. This may be key to evaluating any role these pheromones play in sexual
selection. Be it genetically or environmentally determined, variation in ability to attract
potential mates places pheromone signalling as a prime candidate to be influenced by sexual
selection. Genetic variation in pheromone production certainly exists in the Boll weevil
where selective breeding for its increase has led to a 4.5 fold increase in selected strains as
compared to the parent population (McCoy and Wright, 1990).
What evidence have we already that is compatible with the idea that females use
pheromone production to choose between mates? Male olfactory attractiveness to females is
positively correlated with a male’s subsequent fertilisation success as measured by its P2
value in double matings in the red flour beetle Triboliwn castaneum (Lewis and Austad,
1994). So in this case two possible influences on the pattern of paternity are correlated. In
T. castaneum, “neither males nor females exhibit aggressive behaviour towards members of
their own or the opposite sex”, (Lewis and Austad, 1994) so it is thought that behaviour on
contact contributes little to assortative mating in this species. Assortative mating has been
26
correlated to pheromone characteristics in Ips pini where it is the ratio of enantiomers of

ipsdienol which were found to be correlated to mating success (Teale et al, 1994).
Different females were found to have a consistent preference within the variation of ratios,
and males were found to show highest response to blends similar to those that they were
subsequently found to produce. The evolutionary genetics of these systems warrants further
consideration since signal production and response are not analogous as might be the case
in assortative mating by body size (see discussion in Teale et al, 1994). No correlation was
found between pheromone characteristics and body size. Teale agrees with Herms et al,
(1991) and infers that predation and interspecific competition may be the source of selection
pressures that have resulted in this pattern of pheromone production and response.
Increased reproduction on ‘conditioned’ hosts is often inferred, yet has

surprisingly little documentation. Perhaps a study of life-history traits for individuals reared
on previously infested and uninfested hosts would give an insight into the magnitude of this
possible benefit. It is very difficult to definitively list all hosts that any one species might
utilise. If no collective feeding benefit can be found for an aggregating species on known
hosts, the possibility of significant benefit on undocumented hosts is always possible.
2.13: CONCLUSIONS
To sum up, both sexual and host conditioning functions of aggregation
pheromones are possible given the evidence currently available. It is likely that in those
cases where a conditioning function is implied, both kinds of benefit are occurring for the
signaller. I have looked at evidence for aggregation-pheromone function from the patterns
that can be recorded now, however, it must be remembered that such pheromones may have
initially evolved in a very different context and may have had a different function. For
example, Prostephanus truncatus is currently found as a pest of stored maize and Cassava,
but is also known to be a wood borer (Hodges, 1994), a habit that may have been
influential in determining its use of pheromones. I will be examining the current influence
of the aggregation pheromone signal on mate choice in Prostephanus truncatus by studying
variation between individuals as suggested in the ‘future work’ section of this review.
27
Chapter 3: Basic patterns of signalling
CHAPTER 3: BASIC PATTERNS OF AGGREGATION-

PHEROMONE SIGNALLING AND RESPONSE.
3.1 INTRODUCTION
The individual was chosen as the unit of study in order to investigate the
function of aggregation pheromone in Prostephanus truncatus. It was hoped that by looking
at patterns of signalling and response of individuals, rather than groups of beetles, trade
offs and correlations between individual variation in signalling and other traits could be
investigated. Reliable measurement of the variation in pheromone signal of single insects
was always going to be a difficult, but rewarding challenge. Whilst this work was in
progress, three other bioassays were also designed to monitor the walking response of
LGB to single signalling males (Hodges and Dobson, 1998 in press; Boughton and
Fadamiro, 1996; Scholz, 1997). Previous work used groups of beetles as the unit of study.
Two main methods were employed throughout this work: first the response of
conspecifics was used to detect and quantify pheromone signals; second, pheromone
samples were collected directly onto filters for chemical analysis. The first method, the use
of a bioassay, proved to be more productive although there can be no substitute for direct
chemical analysis to answer some questions. This chapter presents the bioassay design and
some initial experiments designed to map out some basic patterns of pheromone
production. Three aims were addressed: first, to test the bioassay equipment; second, to
obtain knowledge of the influence of some variables on pheromone production and
response, which was required for the design of later experiments; third, to obtain
knowledge needed to elucidate the biological function of aggregation pheromone in LGB.
The main variables tested were: time of day; presentation of maize as whole grains or flour
to the signaller; removal and addition of food to the signaller; type of plant host; and the
presence of females.
The CTH room used in the majority of the laboratory-based experiments in this
study runs on a 12 hours light and 12 hours dark cycle. The light comes on at 8am and
goes off at 8pm (GMT). LGB has already been shown to exhibit a periodicity of flight
activity in the laboratory that is similar to that observed in the field where beetles become
particularly active around dusk and in some cases dawn (Scholz, 1997; Tigar et al., 1993;
Fadamiro and Wyatt, 1995). Ips confusus (Scolytidae) exhibits periodicity of pheromone
production that varies in synchrony with the periodicity of flight activity (Borden, 1967). In
fact there are many such examples of periodicity of pheromone release that coincide with
general increase in activity, which are particularly common for true sex pheromones (see
Burkholder and Ma, 1985, and references therein). Therefore, it was likely that LGB might
28
show a variable bioassay score with time of day. We needed to know whether there were
any times of day when production/response to pheromone was particularly low and
therefore should be avoided in the design of experiments.
Prostephanus truncatus is a tunnelling beetle and it can be difficult to extract

males from their food. Also, when males fall out of grains or are busy tunnelling into a new
grain this may alter their rate of food uptake, which in turn may alter their pheromone
signal. To avoid such complications I tested to see if males will signal if the grain is
presented as a more uniform medium such as coarse ground maize meal.
Information about the timing of initiation and cessation of signalling when

males are put on food or left to starve will address the question of whether male LGB will
produce pheromone when not on food. In other words, do male LGB ever cheat by
signalling in the absence of an appropriate oviposition site or is pheromone production
always an honest signal that indicates the location of an appropriate host? Direct collection
of pheromone from 20 males placed on plaster of Paris chips instead of food showed that
no signal (measured as amount of T l) was being produced after three days (Smith et al.,
1996).
Acoustic signals may be energetically costly, variation in call rate and duration
together account for 80% of the variance in metabolism during calling in the tree frog, Hyla
versicolor (quoted in Andersson, 1994). This has led to the idea that such signals may be
used as an honest indicator of health as well as location. The discussion of whether the
pheromone signal of LGB males is energetically or nutritionally constrained was
investigated further by testing the influence of food type on the signal. Signallers feeding
on the root crop cassava were compared to those feeding on maize. Although LGB is
known to breed on cassava, it is less nutritious than maize and FI offspring are
significantly smaller (R.H. Smith, pers. comm.). Scholz (1997) found no difference in the
attractivity of pheromone signals from males signalling on wood species known to support
LGB reproduction (Commiphora africana ), compared to those feeding on maize even when
both these odour sources were presented simultaneously, although the response to maize
was numerically higher.
Shutdown of production of aggregation pheromone was detectable after five

days exposure of males to females in a Scolytid beetle (Borden, 1967). Pheromone shut
down in LGB can also be induced by the presence of females (Smith et al., 1996)). This
phenomenon is obviously crucial to the interpretation of the biological function of this
signal and was used in later experiments. Smith et al., (1996) found a significant decrease
in pheromone production by males just 24 hours after the introduction of live females. In
Smith et al., (1996), beetles were not allowed sufficient time to reach their maximum rate of
pheromone output before one half of the test insects were exposed to live females. This
29
may reduce the time it takes for the treatments to become significantly different because the
control treatment is still on the increase as the female-treated group declines. I have used my
bioassay to record the timing of shut down of fully signalling males.
3.2 BIOASSAY DESIGN

The bioassay presented here provides information about the net outcome of a
signal and the response to that signal. It can be difficult to ascribe variation in outcome
specifically to the signal or the response. If differentiation is required, artificial sources of
pheromone can be used to provide a reliable signal. Likewise, direct chemical analysis of
the signal could eliminate variation arising from a variable response.
The walking response of beetles moving towards an aggregation pheromone

source was chosen as the variable to be measured in the bioassay. During the initial
development of the apparatus, a failure to see any observable response to a signaller could
have arisen either because the male was not signalling or because the responder was
ignoring the signal, or because the apparatus was not relaying the signal. The use of
artificial pheromone sources to provide a reliable signed was rejected for fear of
contamination with unnaturally high quantities of super attractive chemicals, which might
later mask subtle differences in natural signals. Direct collection of signals to check that
males were signalling was also rejected as an option since this is very time consuming and
has its own technical difficulties.
A wide variety of methods were tested incorporating the use of different shaped
chambers; chambers made of different materials; different aged beetles; beetles isolated
from conspecifics for varying amounts of time; responding beetles kept with or without
food; male and female responders; and different ambient conditions. It was not feasible to
test all possible combinations. Instead, using what was already known about this signalling
system the most likely combinations of insects, conditions and apparatus were tested first.
All apparent success in eliciting some kind of response to the signaller was recorded and the
bioassay was improved step by step until a reliable method was found.
The result was an olfactometer consisting of a choice chamber with four

different air currents leading into a central arena. A reliable signal was produced when
males were kept isolated on food for one week at 30 °C 70% r.h. Females were found to be
more responsive than males to pheromone signals. Responders were also kept isolated on
food for one week prior to use. The number of times the responding beetle placed its head
and prothorax in each of the tubes leading into the central arena within a predetermined
period of time was chosen as a convenient measure of the walking response. To make each
tube choice independent, the responder is placed back into the centre of the arena after each
choice.
30
3.2.1 Details of the apparatus

For a plan view of the bioassay apparatus see fig. 3.2.1a. The plastic petri dish
(central arena) has a hole drilled in the lid connected to a vacuum pump which draws air
from outside the apparatus through the holes in the lids of the source pots and into the arena
through clear plastic tubing. Air currents are thus drawn from four pots containing either
the beetle source to be assayed, or grain only (see fig. 3.2.1b). Grain-only pots act as a
control for the influence of food volatiles and any inherent tendency of beetles to crawl
down tubes. The petri-dish was lined with a disc of paper towel that is very slightly larger
than the base of the dish so it can be made flush with the sides so that beetles are unable to
crawl under it. This provides a rough surface over which beetles can move easily. All
assays were performed in a constant temperature room where temperature, humidity and
light regime were controlled.
Up to three arenas could be assayed together at any one time. Suction was
always adjusted so that each arena received suction at a rate of 1 litre of air per minute.
3.2.2 Procedure during the assay

1. The next sources to be assayed are connected directly to suction tube for 3 minutes to
remove volatiles already present in air in the container.
2. Petri-dish is wiped with 70% ethanol (water solvent).
3. New paper lining is put in the arena.
4. New tubes are used to connect up the pots to the arena (control pots are used several
times).
5. Suction is checked.
6. New source pot is incorporated into the apparatus in a random position around the
arena.
7. A new responder beetle, is placed in the middle of the arena, the lid is put back and the
stop watch is started. All responders were kept in a holding arena similar to the arena of the
choice chamber for a period of one assay prior to testing.
8. The number of times a responder places its head and prothorax into each tube is
recorded for a set length of time (usually 20 minutes). Each time this occurs the responder
is returned to the middle of the arena.
9. Assay is dismantled.
31
FOOD ONLY
Plastic tubing CONTROL POT
(inner diameter 2mm)
To vacuum tap via Responder

flow meter
Signaller
FOOD ONLY
CONTROL POT PHEROMONE
SOURCE POT
Petri dish of diameter 9.5cm and
depth 1.5cm.
Arena lined with a disc of FOOD ONLY
paper towel for easy movement of CONTROL POT
the responder.
Fig. 3.2.1a: Plan view of bioassay apparatus.
Clear plastic tubing of

inner diameter 2mm
connecting container to
arena ----- Plastic lid with additional hole to
allow air into container
— Cylindrical glass pot

Food source Diameter=2.3cm
Height=5cm
Fig. 3.2.1b: Detail of source pots.
32
3.3 MATERIALS AND METHODS

3.3.1 Influence of time of day
Source o f insects:
The males used as pheromone sources were first isolated as pupae. Males were
reared separately on flour for two weeks, sexed and then placed on fresh split grain seven
days before the first assay. Fifty females were used as responders. These were sieved from
a six week old culture, sexed, and then placed on fresh ground grain in a pheromone-free
CTH room (30°C) six days before the first assay.
Apparatus:
Olfactometer apparatus as described in section 3.2. was used for the bioassays.
Each assay was continued for 20 minutes and two arenas were run side by side.
Experimental design:
12 single male pheromone sources were used. Each source was assayed once
during all hours of light in the CTH room. Four replicates were assayed every hour from
1lam-2pm and 5pm-8pm on one day, and then 8am-1lam and 2pm-5pm on the following
day. This sequence was repeated for a new batch of four males every two days. Trials
therefore continued over six days. Responders always had a break of two days between
consecutive trials. Within this constraint responders were randomly allocated to trials.
Analysis:
A Kruskal-Wallis test was used to test for any significant differences of
response with time of day.
3.3.2 Influence of using flour instead of whole grain as food, on the

sig naller
Source o f insects:
16 male signallers were reared from pupae. All males were approximately three
weeks old when they were placed into a treatment group.
Apparatus:
Olfactometer apparatus as described in section 3.2 was used. All assays were
continued for 20 minutes. Two sets of apparatus were run simultaneously.
Males were randomly assigned to one of two food regimes: first,
approximately 3cm3coarse ground maize meal; second, an equivalent amount of split
maize. Males were left in their treatments for 10 days before the first assay. All replicates
were assayed once each per day for four days. One of each treatment was assayed at the
same time.
33
3.3.3 Timing of shut down of signal induced by removal of food

Source o f insects:
20 males of mixed age were used as pheromone sources (taken from a 2 month
old culture). All were placed singly in glass pots with fresh grain and left for 10 days prior
to the experiment. A bank of 40 females was used as responders. The females were all
placed singly in glass pots with fresh grain 10 days prior to the first assay.
Apparatus:
Bioassay apparatus is described in section 3.2. Source pots were adapted to
allow food to be present in a pot, yet not accessible to test insects, see fig. 3.3.3a. Two sets
of apparatus were assayed at once (one from each treatment) and assays were continued for
20 minutes.
Males were randomly allocated to one of two treatments, ‘with food’ and
‘without food’ (see fig. 3.3.3a). Males were placed in test pots and all replicates were
assayed once daily for four days. Replicates were assayed two at a time, one of each
treatment in random order (picked out of a hat).
Analysis:
The response given to each of the treatments at the end of the trials (day 4) was
compared for a difference using a Mann-Whitney U test since there was a small sample size
and the data obtained did not approximate to a normal distribution.
3.3.4 Timing of onset of signalling after placing on grain

Source o f insects:
Male signallers were placed individually on coarse ground grain and left in the
CTH room for one week. All males were then placed in fresh pots containing small pieces
of polystyrene, but no food. Three days later males were assigned to treatment pots. 35
females were used as responders. All were placed individually on fresh, ground maize meal
and left for 10 days before the first assay.
Apparatus:
Bioassay apparatus was used as described in section 3.2. Source pots as
described in experiment 3.3.3 were used (see fig. 3.3.3a).
24 males were randomly assigned to one of three treatments:
1. x8 Kept starved
2. x8 Allowed access to grain
3. x8 Allowed access to grain and one live female.
34
All replicates were assayed once per day on day 0,1,2,4,6,7,9,and 13 after
males were placed into treatment pots. After being assayed on day 13, all males from
treatments 2 and 3 were placed on fresh grain with no females. These pots were then
assayed after a further eight days (day 21).
MALE ALLOWED MALE NOT ALLOWED

ACCESS TO FOOD ACCESS TO FOOD
Hole for aeration
3.2cm
O
c
5.6cm
Lid containing
food -------
O ... through to assay u

Test male
Test male J 1 J ! M L --------- polystyrene pieces----------- BHl
2.3cm
Food
Fig. 3.3.3.a : Cross section of plastic pheromone-source pots adapted to contain food without allowing
insects access to it. NB. LGB cannot walk up smooth vertical surfaces.
3.3.5 Influence of plant host on male-aggregation-pheromone

signal
Source o f insects:
All insects were sieved from a 10 week old culture. All insects were placed
individually on food in glass pots. 50 females were placed on coarse ground maize meal.
48 males were randomly allocated to one of two food regimes:
1. x24 Three split maize grains and a small quantity of coarse maize meal
2. x24 A similar volume of cassava (supplied by NRI see section 1.5.2)
All beetles were placed in the CTH room and left for 14 days before the first
bioassay trials. The quantity of food given was chosen such that it was excess to that which
could possibly be fully tunnelled by the male, yet small enough to limit the possible impact
of differential absorption of pheromone onto the food source on the volatiles released from
the pot as a whole.
35
Apparatus:
Olfactometer apparatus as described in section 3.2 was used. Three replicates
were run at once. Assays were 15 minutes long.
All male sources were assayed singly 14 days after placement in the two food
treatments. The three remaining positions in the olfactometer were occupied by food only
controls. Males placed on cassava had cassava only controls and males placed on maize had
maize only controls. Two replicates of one treatment were assayed at the same time as one
replicate of the other treatment and then vice-versa for each group of three replicates
throughout the day. Within this constraint, the order with which males were assayed was
randomised.
One day later all males were randomly paired up with a male from the other
treatment group. Each pair was assayed together in the olfactometer apparatus in positions
opposite each other. The two remaining positions in the olfactometer were taken up by one
maize-only control pot and one cassava only control pot.
Single females were used as responders and were randomly selected from a
bank of 50. Each female was used a maximum of once per day.
Analysis:
A Wilcoxon signed-ranks test was used to test for a difference between the
responses given to each of the treatments when they were presented simultaneously.
3.3.6 Timing of shut down of pheromone signal induced by Female

Factor.
Source o f insects:
Thirty males were taken from a two month old culture and used as signallers
for this experiment. They were sexed and placed in individual glass pots on fresh maize
meal and left in the CTH room for 10 days. At the end of this period each male was
assigned to one of three treatments:
1. xlO Single male on slightly kibbled grain with some flour, total quantity
equivalent to approximately two whole grains of maize.
2. x 10 As in treatment 1 with a single live adult female.
3. xlO As in treatment 1 but using previously infested grain from the same
culture jar that the signallers were taken from (two month old). This was first
frozen for eight days and then left to acclimate to the CTH room for two days.
36
All females used (as responders and in treatment 2) were taken from a six week
old culture, sexed and placed in individual glass pots on fresh maize meal and left in the
CTH room for 20 days.
Apparatus:
Bioassay apparatus was used as described in section 3.2.
Assays were started the day after placement of males in treatment pots. Assays
were performed daily for four days and then every other day until 10 days of treatment had
passed.
Each assay was 20 minutes long and three sets of apparatus were run
simultaneously, one allocated to each of the treatments randomly. Responders were selected
randomly from a bank of approximately 40 and no responder was used more than once on
the same day. Control pots contained the same grain as that used in treatment 1.
Analysis:
Mann-Whitney U tests were used to compare separately both Female-Factor
treatments with the males signalling on fresh grain.
3.4 RESULTS
3.4.1 Influence of time of day
The frequency distribution of responses recorded is skewed to the left since
many assays recorded low or zero responses. The data are therefore represented as medians
and their interquartile ranges.
The overall median response per trial given to source pots (those containing a
male) was 3 visits (25% quartile=l, 75% quartile=6). The overall median response per trial
given to food only control pots was 0 visits (25% quartile=0, 75% quartile=0.333).
Control pots elicited a significantly lower response than source pots (Mann-Whitney U test
stat. = 2337, N=144, p<0.0001). The response to single male pheromone sources was
found to be higher than that for control tubes for all hours of the light phase of the CTH
room (fig. 3.4.1). Therefore source pots retain some attractivity throughout this time and
females are responsive to this signal throughout this time. No significant pattern of
changing response with time of day was detected: Kruskal-Wallis test statistic = 6.62,
d.f.=l 1, p=0.83.
37
a Single male on food

□ Food only control
1 2 -i --------------------------------------------
Time since the light switched on in the CTH room (hours)
Fig. 3.4.1: Bar chart to show the median response given to single males placed on food for all hours of the
light phase in the CTH room. Error bars span the 25% and 75% quartile values. N=12 for all columns. No
significant pattern of changing response with time of day was detected: Kruskal-Wallis test statistic = 6.62,
d .f.= ll, p=0.83.
3.4.2 Influence of using flour instead of whole grain as food, on the

s ig n a ller
N o significant difference was found in the average response to signallers
placed on w hole or ground grain: whole grain median = 7 visits per trial (25% quartile = 4;
75% quartile = 8.25); ground grain median = 6.5 visits per trial (25% quartile =1.75; 75%
quartile = 9) (N =8 for each treatment).
3.4.3 Timing of shut down of signal induced by removal of food

The median and inter-quartile range o f response to signallers from both
treatments was initially o f comparable value. The median response recorded for signallers
deprived o f food consistently dropped each day until reaching zero by day three. All
signallers from the food-deprived treatment were still alive up until day four, but two had
died by day six, the rest had died by day eight. All signallers given food lived until the end
o f the experiment. On day four (the last day that all subjects were alive), m ales on food
were significantly more attractive than those allowed to starve (Mann-Whitney U test
stat.=84.5, ld .f., p < 0.001) (see fig. 3.4.3). The response given to all food only control
pots remained low throughout the experiment (overall median = 0; 25% quartile = 0, 75%
quartile = 0.333).
38
7
(1) Male with food
6 (2) Male with no food
03
5
CD
Q-
cfl 4
U
cCn
oCX 3
000>
CO 2
<D
>
<
1
0
0 1 2 3 4
Time since start of treatment (days)
Fig. 3.4.3: Median response with time to single males either on food or without food compared to food
only control pots. Error bars span the 25% and 75% quartile values. N=10 for each point.
3.4.4 Timing of onset of signalling after placing on grain

A detectable response to signallers placed on food with no female (compared to
control pots) was recorded after six days and was maintained until the end of the
experiment (day 21) (see fig. 3.4.4). The average response from day 6 until day 13 was
fairly low at around one and a half visits per assay. Since each daily estimate has a fairly
large variation, it was possible that the treatment difference could have arisen from just a
few individuals. However, only two of eight replicates elicited a consistently low response
and the other six contributed about equally to this result.
No detectable response was ever recorded for signallers kept without food and
all replicates of this treatment were dead by day three (this data is not shown in fig. 3.4.4).
No detectable response was recorded for signallers placed on food with one
live female, for 13 days. These signallers were shown to be capable of signalling since a
response was recorded on day 21, eight days after the females were removed and the
signallers were placed on fresh grain (see fig. 3.4.4). This response was comparable to that
recorded for the males on food only treatment.
39
♦ — Male on food
7 ♦ — Male on food with 1 female
— - Food only control (all treatments
combined)
6
03
5
<
a.D FEMALE
REMOVED
4
<u
co
a, 3
<u
00
ca. 2
U
<0
>
<
1
0
0 5 10 15 20 25
Time since placed in treatment (days)
Fig 3.4.4: Median response against time to single-male pheromone sources previously starved and then
placed in one of two treatments: placed on food; placed on food with one live female. The arrow indicates
when all females were removed and all sources were placed on fresh grain. Error bars span the 25% and 75%
quartile values. N=8 for each point.
3.4.5 Influence of plant host on male-aggregation-pheromone

signal
Males assayed singly:
The median response to male sources was 5 visits per 15 minute assay (25%
quartile=3.75,75% quartile=8). As in other bioassay trials, response to the control pots
was consistently low (median = 0). The median responses recorded for males feeding on
maize or cassava were identical, although the mean response to cassava was marginally
smaller (see fig. 3.4.5a). Cassava and maize only controls elicited a comparably low
response compared to pots containing beetles (see fig. 3.4.5a).
Males assayed in pairs:

The median response given by the female responder to either of the two male
sources was 6.5 visits per 15 minute assay (25% quartile=4, 75% quartile=9). Again
response to the controls was consistently low (median = 0). In these trials males feeding on
maize were found to be significantly more attractive to those feeding on cassava: Wilcoxon
40
signed-ranks test, z= -2 .1 4 4 , N = 24, p=0.032. The median response given to m ales feeding
on cassava w as just 50% that given to males on maize (see fig. 3.4.5b).
£on
<
D-
Q
C
o
&
T T
Maize Maize Cassava Cassava
source control source control
Fig. 3.4.5a: The median response to single males signalling on either maize or cassava compared to food
only control pots. Males were assayed SINGLY i.e. one male pheromone source per assay. Error bars span
the 25% and 75% quartile values. N=24 for each column. The males feeding on cassava elicited a
numerically lower, but not significantly different response from males feeding on maize.
0
Maize Maize Cassava Cassava
source control source control
Fig. 3.4.5b: The mean response to single males signalling on either maize or cassava compared to food
only control pots. Males were assayed IN PAIRS i.e. one male signalling on maize was placed opposite
41
one male signalling on cassava in each assay. Error bars span the 25% and 75% quartile values. N=24 for
each column. Males feeding on maize were found to be significantly more attractive to those feeding on
cassava: Wilcoxon signed-ranks test, z=-2.144, N=24, p=0.032.
3.4.6 Timing of shut down of pheromone signal induced by Female
Factor.
Response to signallers was comparable for all treatments for the first four days
of trials and wavered around a response of four visits per trial. However response to both
Female Factor treatments became significantly lower than the control line on day six and
generally remained low until the end of the experiment. The significance of this difference
is indicated on fig. 3.4.6.
—o—Single male on fresh grain

—• —Single male plus live female
—A—Single male on previously infested grain
5
4.5
4
8o
8 3.5
£ 3
CO
o 2.5
j5
<u
SP 2
<5
> 1. 5
<
1
0.5
0
0 2 4 6 8 10
Time (days)
Fig. 3.4.6: Graph to show how pheromone shut down in response to Female Factor varies with time since
treatment started. Bioassay score is the number of visits per assay. The average (shown as the median)
scores given to males placed in three different treatments are shown: males placed on fresh grain; males
placed on fresh grain with one live female; males placed on previously infested grain. Males are potentially
exposed to Female Factor in the last two treatments. N=10 for each point. Points marked as ‘a’ are not
significantly different from the points on the male placed on fresh-grain line (Mann-Whitney U test
p>0.05). Points marked as ‘b’ are significantly different from the males on fresh-grain line (Mann-Whitney
U test p<0.05).
3.5 DISCUSSION
Time o f day:
Since a response to male signallers has been found throughout the light phase
of the CTH room, all future experiments can appropriately be performed during this time.
42
No significant influence of time of day was found on the response level in the bioassay, but
this potential influence will continue to be controlled for when considering treatment
effects. The grain in pheromone-source pots may act as a sink for pheromone and thus
buffer any daily fluctuation in production such that responders are unlikely to show a
response that is linearly proportional to the size of the signal.
Effect o f male starvation:

Detectable changes in pheromone signal on removal of food occur on a similar
time scale to starvation, i.e. of the order of a few days as opposed to a few hours. This
shows that it is possible for males to emit some signal when there is no food available.
Pheromone emitted during the first two days after food was removed may have been
synthesised before this time. Also, whether signals produced without food are strong
enough to be attractive in a natural situation cannot be determined using this apparatus. As
already noted this apparatus may be relatively insensitive to changes in signal size down to
a threshold level.
The resumption of detectable signalling after a period of starvation was

recorded six days after the reintroduction of food. This long time lag implies that the signal
may be energetically costly and/or slow to synthesise. Males kept in the presence of females
did not resume signalling for the total of 13 days, which agrees with the findings of Smith
et al, (1996) that females can suppress male signalling for at least 15 days. It would be
interesting to see if males whose signal is repressed by the presence of a female resume
signalling faster than those repressed by starvation.
Effect o f different plant hosts:

The fact that males signalling on a less nutritious host plant were less attractive
supports the hypothesis that the aggregation pheromone signal is costly to produce, either
energetically or nutritionally. If the signal were cheap to produce then we would predict that
males would either give a maximum signal or no signal at all. There is no reason to suppose
that males would willingly inform potential responders of the lower quality of the host.
Thus, males may only be able to produce the most effective signals when on the most
nutritious plant hosts. If this is indeed the case then aggregation-pheromone signals may
play a key role in host-plant selection, particularly when many potential hosts of different
quality are available and responders are able to compare signals.
This experiment demonstrates the increased sensitivity of the bioassay to

differences in two signals that can be gained by presenting two signals simultaneously to a
responder. This increased sensitivity has two sources: first, the behaviour of the responder
and second, the creation of paired data that can eliminate the effect of differences in overall
response by different responders. Later experiments have used this method to detect
differences between pheromone signals.
43
Effect o f Female Factor:

The time-scale of detectable pheromone shut down recorded in response to the
presence of females is approximately twice as long as that found in response to starvation.
The start of detectable shut down was also at least six times slower than that recorded by
direct chemical analysis in Smith et al., (1996). A more sensitive method of recording shut
down would be to feed the signals from treated and untreated males into the same arena as
described in experiment 3.3.5. In later experiments where Female Factor was used to
manipulate the pheromone signal, males were exposed for a period of at least seven days to
ensure significant pheromone shut down.
44
Chapter 4: Inter-male variation (I bioassay and II chemical analysis)
CHAPTER 4: INTER-MALE VARIATION IN PHEROMONE

SIGNAL (I BIOASSAY and II CHEMICAL ANALYSIS)
4.1 INTRODUCTION
The main aim of the work presented in this chapter was to test if variation in
male-pheromone signals translates into a perceptible difference in the response elicited in
conspecifics. As Andersson says, “For demonstration of sexual selection of a trait, it is
important to show that variation in the trait leads to variation in a possible mechanism of
sexual selection”. It is already known that aggregation-pheromone characteristics vary with
age, level of starvation, and in response to other pheromone signals in Prostephanus
truncatus (Smith et al., 1996 and chapter 3 this study). Direct collection of pheromone
from groups of 10 males showed that pheromone production increased rapidly with adult
age to a peak of production at 10-15 days. After this peak, production declined steadily
until death (Smith et al., 1996).
If the pheromone signal is a sexually selected trait then experience of other such
systems leads us to predict that it will vary between individuals and such variation is often
heritable (see Cade, 1984, and Andersson, 1994). Linkage disequilibrium, shifting
selection pressures, the existence of many sexually selected traits and the possibility of
alternative strategies may also tend to increase heritable variation (Cade, 1984). In
Prostephanus truncatus signalling can lead to increased predation of larvae by Teretriosoma
nigrescens, which uses the signal as a kairomone to locate its prey. Increased risk of
predation is one possible cost of signalling that may influence the evolution of the signal.
The large potential for changing population density arising from the formation of
aggregations may represent an influential shifting selection pressure on the relative benefits
of signalling.
The question of whether a suite of characters is sexually selected in LGB will

be investigated in later chapters. Since males can be signallers or responders there is the
possibility that two or more strategies can be favoured for different males, or the same male
under different conditions or of different ages. One particularly interesting idea was
suggested by Moore (1995), that, “signalling by a male may be beneficial only if he is the
most attractive male present. Furthermore, signalling by a subordinate male may indicate
his presence to the dominant male. Therefore the tactic of less attractive males should be to
remain silent”. This mirrors the situation found in Cowbirds (Monothrus species) (West,
1981 quoted in Moore, 1995), where only dominant Cowbirds sing. These processes
would tend to increase the observed phenotypic variation in signalling.
45
Chapter 4: Inter-male variation (/ bioassay and II chemical analysis)
The temporal stability of any hierarchy of attractiveness between males will

indicate the relative importance of the identity of the signaller in determining pheromone
characteristics. Natural variation in pheromone characteristics, be it a stable characteristic of
the individual or more facultative in nature, presents an opportunity to evaluate the
consequences of signalling. This opportunity is exploited in chapters 5 and 6.
The main challenge of this work was the technical difficulty of measuring
olfactory signals. In this work we have not measured chemical output directly, instead it
has been implied from behavioural responses. Measuring the pheromone signal indirectly in
this fashion increases the likelihood of error of measuring. Many variables, such as the
activity level of the responder, have been controlled for by using a contest-type method
where two male signallers are simultaneously presented to a responder. Still, any
preferences shown only gain credence if they are shown to be repeatable. If the variation
fluctuates greatly in time then it is difficult to distinguish real fluctuating pheromone signals
with large random components in the method of pheromone measurement.
Three main sets of trials designed to assess the predictability of response to

individual variation in pheromone signal are presented. First, likelihood of inter-male
variation was maximised by testing males originally taken from different geographic
locations. Second, trials designed to compare the response of male and female responders
were conducted. Fadamiro found that there was no difference in the response of males and
females to an artificial pheromone source in a flight bioassay (Fadamiro, 1995), or to an
artificial pheromone source in a walking bioassay (Boughton and Fadamiro, 1996).
Hodges and Dobson (in press) and Scholz (1997), however, found female response to be
higher than male response in walking bioassays. The discrepancy may be because insects in
the trials conducted by Boughton and Fadamiro were not isolated from culture prior to
testing, whereas those in the other studies were isolated prior to testing. If the sexes are
isolated prior to testing then there could be differences in their culture conditions. More
specifically, males may be exposed to their own aggregation pheromone signal, whereas
females do not produce such a signal. If beetles can become habituated to the signal, then
males may exhibit a lower response in the bioassay. Finally, we tested for the existence of
detectable variation between male signals when age, strain, ambient conditions and
culturing regime of the signallers were all kept constant. The temporal stability of this
variation was also investigated.
Age of test insects was deliberately not always constrained to be of uniform age
to get an idea of the range and pattern of variation in signal and response across real
populations. Very young insects (less than 5 days old as adults) were not used since both
males and females are particularly inactive and males do not produce a detectable
aggregation pheromone signal at this age (Smith et al, 1996). Previous work has indicated
that there is no clear switch in male strategies (to be a signaller OR responder) with age
46
(Smith et al., 1996), however, non-continuous variation in signalling could be detected by

using a range of ages. No such dichotomy was found. This study focuses on the
consequences of the variation in male signals on reproductive success. Future work could
particularly address changes in behaviour with age of insects. Obtaining known age insects
of LGB is hampered to a certain extent by the difficulty in removing parent generations
from their tunnels within grains. Scholz (1997) has developed a method of using
compacted flour (which gives better insect survivorship than loose flour, yet is easy to
sieve) as a culture media which would make such a study more practical.
An attempt was made to correlate characteristics of the pheromone signal

(absolute amounts of T1 and T2 and their ratio T1/T2) with signal success in the bioassay.
This was made possible by collaborating with an experiment performed at NRI Chatham
where pheromone signals from single males were collected, and whose chemical
components were subsequently quantified. Bioassay-choice trials were conducted on these
males immediately after the last chemical sample was taken. It was hoped that this would
show which features of the pheromone signal are important for short range attraction such
as that measured in the bioassay. Also, the sensitivity of the bioassay could be chemically
quantified.
4.2 AIMS
• Find out if variation between signalling males was detectable using the bioassay.
• Find out if responders ‘agree’ as to which males are most attractive.
• Find out if response to male signallers differs between the sexes in overall magnitude
and in the pattern of preference between signallers.
• Find out how stable any hierarchy between male signallers is over time.
• Correlate chemical characteristics of pheromone signals to their performance in the

bioassay.

4.3.1 Experiment to determine if variation in male signal was
detectable using the bioassay.
This experiment uses signallers of different strains to maximise the likelihood
that there will variation in the aggregation pheromone signal. Only females were used as
responders in this early experiment since they gave consistently higher responses than male
responders in preliminary trials.
47
Chapter 4: lnter-male variation (I bioassay and H chemical analysis)
Source o f insects:
Six male signallers were tested. Two each from a Mexican, Togo and
Tanzanian source were used. All males were removed from cultures as enclosed pupae.
Hatching date of each insect was noted and fresh adults were placed individually on ground
grain. These insects were cultured for a week before sexing to minimise damage. After
sexing they were all placed singly on fresh grain and left for five days before the first trial
(all males approximately aged 12 days as adults at first trial).
30 females taken from a Tanzanian source were used as responders. These

responders were adults randomly sieved from a culture, sexed and then placed individually
in glass pots with some ground grain. Responders were isolated from conspecifics in the
CTH room for five days before the first trial.
Apparatus:
All six males were screened by assaying singly at age 12 days and age 18 days.
This was done to check that males were signalling enough to elicit a response in the
bioassay. Males (aged 19-20 days) were then placed two at a time into the assay apparatus
such that they occupied positions opposite each other. Every possible pairing of the six
males was assayed each day for two days. Each day the assays were performed in a
random order (picked out of a hat) within the constraint of having two sets of apparatus
running at once. Each assay was continued for 40 minutes and two sets of apparatus were
run simultaneously. All source pots were aerated for approximately 4 minutes before each
assay to reduce any influence of the timing of previous tests on pheromone levels.
Analysis:
We wanted to know if the outcome of a contest between two males is
predictable from how those two males have performed in other trials. To achieve this the
outcome of any given contest (score given to male 1- score given to male 2) was plotted
against the difference in overall scores obtained by the two males excluding data from
the contest under consideration as recorded on that day’s trials:
Difference in overall scores = ((total score given to male 1- total score

obtained by his opponents) - (total score given to male 2 - total score obtained by his
opponents)).
Males are referred to as male 1 or male 2 in the results section. The rule for
calling a male as 1 or 2 is as follows. The males were placed in an arbitrary order of:
Tanzania 1, Tanzania 2, Togo 1, Togo 2, Mexican 1 and Mexican 2. In any contest
48
Chapter 4: Inter-nuile variation (I bioassay and II chemical analysis)
considered, the male highest up this rank was called male 1 and the lower ranking male,
male 2.
Lillifors analysis (in SYSTAT) (Wilkinson, 1990) showed the data sets in this
chapter to be approximately normally distributed. All correlation coefficients presented are
Pearson correlation coefficients calculated using Excel 5.
4.3.2 Experiment to determine if males and females 'agree’ which

of a pair of males is more attractive.
In this experiment, both males and females were used as responders to allow a
direct comparison to be made between the overall level of response of both sexes and any
difference in preferences shown for particular signals. This experiment focused on one
geographical source of insects (Tanzanian source). Choice of strain was arbitrary.
Source o f insects:
Nine males were tested. They were sieved from culture as adults of unknown
age and placed on fresh ground maize for nine days before the first trial. All males used
were from a Tanzanian source. Thirty females were isolated per day for three days from a
Tanzanian source. All were sexed and placed directly on fresh ground grain and were
isolated from culture two days before being used in the bioassay.
Apparatus:
All combinations of male pairings were tested twice, once with a male
2
responder and once with a female responder (total number of assays = ((9 -9)/2)x2=72).
Assays were run on two sets of apparatus simultaneously with one set using a female as a
responder and the other a male. Within this constraint assays were performed in a
randomised order (picked out of a hat) over three days. Each assay was continued for 25
minutes. All source pots were aerated for approximately four minutes before each assay to
reduce any influence of the timing of previous tests on pheromone levels.
Analysis:
Results using male responders were processed separately from those results
using female responders. Data were processed as described in experiment 4.3.1. The
preferences of the sexes were compared by correlating the outcome of each signal
combination as given by a female responder vs. the equivalent outcome given by a male
responder. Finally the average magnitude of response irrespective of preference, of each of
the sexes was compared using a Wilcoxon signed-ranks test.
49
4.3.3 Experiment to determine if the response of females to

aggregation pheromone can be depressed by increasing their
exposure to the signal prior to testing.
Source o f insects:
All insects were sieved from the same 10-week old culture and sexed. 25 males
were placed in separate glass pots with a single pre-tunnelled whole grain and a small
amount of maize meal. 50 females were placed in similar pots with approximately 3mm
depth of maize meal. All pots were covered with a small piece of nylon mesh (100 holes
per cm2) held in place with tape. The shallow depth of food ensured that females were
continually exposed to the air circulating around the pot.
Apparatus:
The bioassay apparatus as described in section 3.2 was used.
Pots containing females were randomly allocated to one of two treatments:
exposure to high or low concentrations of aggregation pheromone. Females allocated to the
low-pheromone-concentration treatment were placed five at a time in a 550ml jar with five
pots containing a small amount of maize only. Females allocated to the high-pheromone-
concentration treatment were placed five at a time in a 550ml jar with five-male-pheromone-
source pots. All jars were sealed with a filter-paper top as described in section 1.5.3. It was
hoped that this method would allow air to diffuse between pots within each jar easily and
thus expose the females in the high-pheromone treatment with air from the signalling males
within their jar.
Insects were left in their treatments for eight days before bioassay trials were
conducted. All trials were conducted during one day. Replicates were assayed three at a
time, one from one treatment and two of the other. Male sources were taken straight from
the high-pheromone-treatment jars. This was done such that females were always taken out
of jars before males so the treatment was maintained right up until each bioassay trial. Male
sources were all assayed twice, once by females from each of the two treatments. Male
sources may have given off slightly less pheromone the second time they were used as any
reservoir of pheromone trapped in the pot was evacuated. Therefore, the order in which
males were presented to each of the two treatments was controlled so each treatment was
exposed to the same number of first and second time used males. All trials were continued
for 20 minutes.
Analysis:
Male identity accounted for a portion of the variation and each male had been
tested with one female from each treatment, therefore the data were analysed using a
Wilcoxon signed-ranks test between females exposed to the same male.
50
4.3.4 Experiment to determine if variation exists between males of

the same source and age, and if so, whether the pattern of variation
is stable over time.
Source o f insects:
Thirty six males were used as signallers and 36 females were used as
responders in this experiment. All were sieved from a Tanzanian culture as pupae and
separated into individual glass pots of coarse ground maize flour. Date of emergence was
noted to obtain adults of known age. Males were weighed and placed into pairs of similar
emergence date, (kept to within at least two days). Weight was noted to allow a preliminary
assessment of the correlation of weight with pheromone attractivity. All beetles were
cultured singly for 10-14 days before the first bioassay trials.
Apparatus:
Bioassay was used as described in section 3.2.
Bioassays were of contest type. Two males were connected up to the arena
simultaneously as described in experiment 4.3.1. Eighteen pairs of males were tested. Each
pair was assayed using one female responder on one day and then a different female
responder on the next day (day 2). All males were then placed on fresh grain and this
procedure was then repeated after a day break (on days 4 and 5) using the same females
with the same male pairing. Males were placed on fresh grain to eliminate the possible
influence of a slowly released pheromone reservoir coming from the grain around the
signaller. Two comparisons were therefore made possible: first, the response of one female
to a pair of males compared to the response of another female to that pair of males the next
day; second, between the response of a female to a pair of males and her response to that
pair after they have been placed in a fresh food tube and after 3 days have elapsed. Trials
were staggered such that the entire experiment was conducted over 7 days. Each female
responder was never used more than once in a day and each pair of males was tested a
maximum of once per day. Replicates were assayed three at a time and each assay was
continued for 25 minutes.
Analysis:
The outcome of all trials was calculated as in experiment 4.3.1. All associations
were tested using a Pearson correlation coefficient (Excel 5).
4.3.5 Experiment to correlate bioassay response to chemical assay,

for pheromone plumes from single males.
Source o f insects:
The beetles used in this experiment were a Ghanaian strain collected from the
field in 1996. They had been maintained at NRI Chatham in a CTH room at 27°C and 70%
r.h. Males were removed from their pupal cases and placed on wheat flour for two days
51
(sexed according to Shires and McCarthy, 1976). Males were then placed in a pre-drilled
tunnel in a maize grain.
Forty females were isolated as adults (sexed as described in section 1.5.6).

These insects were used as responders. They were sexed then placed on ground maize
grain in separate glass pots in a CTH room set at 27°C and 70% r.h. and left for 16 days
before the trials.
Bioassay apparatus:
Bioassay apparatus as described in section 3.2 was used. Three replicates were
assayed at once. A pump was used instead of the vacuum line, this was maintained at about
4.5 litres per minute ensuring that each set of apparatus received an air flow of
approximately 1.5 litres per minute.
Pheromone collection apparatus:

Collection filters were made using Pasteur pipettes, packed with 200mg of
porapak-Q (Phase Separations, UK), and plugged at both ends with silanised glass wool.
The filters were prepared for use by flushing with ‘distoF grade dichloromethane (Fisher
Scientific, UK). Beetles in maize grains and control grains without beetles were placed
singly in 30 cm3glass vessels (Fisher Scientific, UK) through which air was drawn by
small electrical pumps at a rate of 1 litre per minute. A large glass round bottomed flask
(1litre capacity)was also connected to the system to act as a buffer against pressure
variations induced by the pump. The intake air was passed through filters containing
200mg of ‘Porapak Q’ (Phase Separations, UK) to collect T1 and T2 (all designed and
performed by staff at NRI).
Chemical analysis:
Volatiles were desorbed from the filters by washing with 750j l l 1 of ‘distol’
grade dichloromethane (in three aliquots). 5jig of octylacetate was added to each sample as
an internal standard. Samples were analysed by capillary gas chromatography using a CP-
Wax52-CB 25m x 0.32mm column (Chrompack, The Netherlands) with helium carrier
gas, at an inlet pressure of 5psi. The temperature was held at 50°C for 2 minutes, then
programmed to 220°C at 6°C/min. Results were calibrated against known amounts of pure
synthetic pheromone, and peak identities were confirmed using an ion-trap detector.
Initially, volatiles were collected at intervals of 1,2 or 3 days for a total of 24
days. After this an additional collection was then made over 60 hours ending the on the
morning of the bioassay.
Males were bioassayed in pairs. Males were initially divided into two groups:
replicates 1-5; and replicates 6-10. All possible combinations of pairs within each group
52
were assayed. In addition to these tests, 10 other pairings were tested such that a male was
taken from each group and all males were tested twice. All of these combinations were
assayed in random order within the constraint that any male could only be chosen once for
each set of three trials to be performed simultaneously. Only one day was available for
trials.
Assays were continued for 20 minutes each. Female responders were only
used once.
4.4 RESULTS
4.4.1 Experiment to determine if variation in male signal was
detectable using the bioassay.
Trials where males were assayed singly:
In the first trial (males aged 12 days) no response was observed for any of the
sources with the exception of one male of Mexican origin who obtained 6 visits per assay.
In the second trial (males aged 18 days) a response was observed for 5 out of 6 of the
sources. The average score obtained was 8.5 visits per assay. Therefore, response to the
signallers was high enough to begin the choice trials.
Trials where two male sources were assayed together:

Mexican 2 escaped on the second day of the trials. The overall score of the
remaining males predicted the winner of any one single trial in 21 of 22 cases where the
responder showed a preference. In other words females ‘agree’ which males are more
attractive and this ranking is stable over the period of each set of trials (8 hours). In addition
to this, the magnitude of the choice (size of the outcome) was correlated to the magnitude
of the overall score difference between the pair of males, (^=0.66, N=25, p<0.001)
(see fig. 4.4.1).
When males were ranked in terms of their ability to attract females the two
males taken from a Tanzanian source occupied the top two ranks. (NB The responders
were also of Tanzanian origin).
53
12
10
8
• •
6
5 -40 30 20 -10 10 20 30 40 50
- 2
-4
»
- 6 H
- 8
Overall score difference
Fig. 4.4.1: Scatter plot to show the correlation between the outcome of a single trial between two males
and their overall score difference as determined by all other trials they have taken part in (^=0.66, N=25,
p<0.001). NB. Females used as responders.
4.4.2 Experiment to determine if males and females ‘agree’ which

of a pair of males is more attractive.
The overall performance of each signaller predicted the winner of any one
single contest in 48 of 62 cases where the responder showed a preference. This shows that
within each sex there is some agreement as to which males are more attractive. This level of
predictability was exactly the same whether male or female responders were used. As in
results section 4.4.1, the magnitude of the preference was correlated with the magnitude of
the overall score difference between the pair of males (for male responders ri=0.37, N=36,
p<0.001, for female responders rMD.26, N=36, p<0.001) (see figs. 4.4.2a and 4.4.2b).
However, it can be seen that the association between these variables is not as strong as that
found in fig. 4.4.1.
The association between the outcome of a trial with a female responder vs. the
outcome of an equivalent trial using a male responder was investigated. These two variables
were found to be significantly, yet loosely correlated (r2=0.15, N=36, p<0.025) (see fig.
4.4.2c). Therefore males and females ‘agree’ which of a pair of signalling males is more
attractive. Females showed a higher level of total response than males with an average of
8.44 visits per trial compared to 4.76 visits per trial when male responders were used. The
total response (male 1 + male 2) given by a female was compared to the equivalent response
54
Chapter 4: Inter-male variation (I bioassay and U chemical analysis)
given by a male using a Wilcoxon signed-ranks test Females consistently gave a higher
total response than males: z=-3.437, N=36, pcO.OOl.
8 0 -]
60 -
<
ou I
c 40 i
2
,p
%-< 20j ! « h § i i
'•3
p ,—
0
y/5 - 11 w
C 0 • l 10 15 20
• 0- 2 0
1s
> -4 0 -
o
-6 0 -
-8 0 -
Outcome of single trial

Fig. 4.4.2a: Scatter plot to show the correlation between the outcome of a single bioassay between two
males and their overall score difference as determined by all other bioassays they have taken part in as
determined by male responders (r=0.37, N=36, pcO.OOl).
80 -
•
60 - •
uo
c
2 • •
:-t—i
20 - • • •
•
'•£
2c
%
«tt :
1
-40
•
p - t o #
10 15 20
>
O - 4 0 •i
•• -6 0 -
-8 0
Outcome of single trial
Fig. 4.4.2b: Scatter plot to show the correlation between the outcome of a single bioassay between two
males and their overall score difference as determined by all other bioassays they have taken part in as
determined by female responders (r=0.26, N=36, pcO.OOl).
55
10 •
S-H 8
T<3D
G
o
o- 6 • •
C/3
e •4
•9
X>
c
<D 1
> -10 ••o 10 20
W) » -2 i
I • » . 4 -{
o
a3
o - 6
- 8
Outcome given by female responders
Fig. 4.4.2c: Scatter graph to show the correlation between the outcome of a single bioassay as determined
by a male responder and the outcome of the equivalent trial using a female responder (FsO.lS, N=36,
p<0.025).
4.4.3 Experiment to determine if the response of females to

aggregation pheromone can be depressed by increasing their
exposure to the signal prior to testing.
Females previously exposed to higher levels of aggregation pheromone were
found to show approximately half the level of response to an aggregation-pheromone signal
in a walking bioassay compared to females kept in low-aggregation-pheromone
concentration conditions (average response for high pheromone treatment = 1.71 ± SE 0.46
visits per trial; average response for low pheromone treatment = 3.52±SE 0.74 visits per
trial). This treatment difference was significant, Wilcoxon signed-ranks test: z=2.523,
N=21, p=0.012.
4.4.4 Experiment to determine if variation exists between males of

the same source and age, and if so, whether the pattern of variation
is stable over time.
For any one pair of males, the outcome given by one female on one day
(response to male 1 minus response to male 2) was correlated to the outcome given by
another female on the next day, (r2=0.47, N=36, pcO.OOl) (see fig. 4.4.4a). Therefore
variation exists between males of the same age and strain that is perceived by female
responders. Also these responders agree which males are more attractive and this hierarchy
is stable over 24 hours.
The response of one female on one day is not significantly correlated to her
response after three days and after the males had been placed on fresh grain, (^=0.034,
N=36, p>0.05) (see fig. 4.4.4b). The ranking of male attractivity is not therefore,
56
Chapter 4: Inter-male variation (I bioassay and // chemical analysis)
completely stable and can be disrupted either by time, disturbance, or a combination of the
two.
10
+ 8 -
IX
eo
T3
6 -
C
o
CQ 4
JV • •
13 2 • • •
6
<22
X>
c -1 5 -1 0 - 5 • §i 10
o> 15
'5b - «•
CD
B
oCD
3
o - 6 -
- 8
Outcome given by female A on day Y
Fig. 4.4.4a: The correlation between the outcome (response given to male 1 minus response given to male
2) of a trial as determined by one female on one day vs. the outcome of the equivalent trial as determined by
a second female on the next day (d=0A7, N=36, pcO.OOl).
15 •
CO
+
>-> 10 •
as
T3
G
O
5 •
.2
"52 • •
B
<22 r -
£ - 8 - 2 ? •
c<D
> - 5
'5b
(D
B
o
o 10
S3
o
-15
Outcome given by female A on day Y
Fig. 4.4.4b: The correlation between the outcome of a trial as determined by one female on one day vs. the
outcome of the equivalent trial as determined by the same female three days later after the male signallers
had been placed on fresh grain (r2=0.034, N=36, p>0.05).
57
Chapter 4: inter-mule variation (1 bioassay and II chemical analysis)
Larger males did perform better in bioassay trials, though this difference was
not statistically significant in these trials. Each pair of males was tested four times in these
trials. Which of a pair of males won the most of these four contests was noted (draws were
also recorded). These frequencies are shown below in table 4.4.4c.
Table. 4.4.4.c: Table to show which of a pair of males won most contests (out of the four performed by
each pair), the larger male, the smaller male or an equal number of contests by each male.
1 Larger male won most Smaller male won most Each male won same
s rnntftsfs cnntftstc nnmhftr o f rnntftsts
4.4.5 Experiment to correlate bioassay response to chemical assay of

pheromone plumes from single males.
Chemical analysis:
Results from the first 24 days of pheromone collection will be reported by
NRI. Here, only the last pheromone collection is reported.
Six of the ten males were found to be emitting detectable amounts of T1 and
T2. Of these six males the average amounts of T1 were 6.15|ig (0.103fig per hour) and the
average amounts of T2 collected were 2.34|ig (0.039|ig per hour). The rate of T1 emission
is plotted against the rate of T2 emission for each signalling beetle in fig. 4.4.5a to
demonstrate the variability in the sample used. Both the amount of each component and the
ratio between them is variable between individuals.
0.07 -
0.06 -
co G
om
0.05
'c/J u,
12 0.04 H
« u.
£ 8.
0.03
S 00
0.02 -
o2 0 . 01 -
'5
0*
0 0.05 0.1 0.15
Rate of T1 emission (micrograros per hour per insect)
Fig. 4.4.5a: Scatter graph to show the rate of T1 emission against the rate of T2 emission for each
signalling beetle.
Bioassay results:
58
The average response given to males that were subsequently found to have
been emitting a detectable amount of either T1 or T2 was 1.64 visits per trial (N=36). The
four males with no detectable signal from the chemical analysis obtained very few visits
during the bioassay trials (average 0.13 visits per trial (N=24). Therefore, both bioassay
and chemical methods have detected the same males as non-signallers. Further
interpretation of the data is now limited by the low number of trials featuring two signalling
males. Two signals are needed to compare the relative success of different features of the
pheromone chemistry. Features of the bioassay-winners’ pheromone chemistry are
presented below in figs. 4.4.5b-d. Where no preference was shown in the bioassay and/or
no difference was found between the pheromone output, the trial is scored as being neutral.
The ratio of the two components of aggregation pheromone (mass of T 1/mass of T2) can
only be calculated for males with a detectable signal.
Table 4.4.5b: Frequencies of winners of bioassay trials in terms of relative amounts of T1 detected in their
pheromone signal. Trials where the males drew in the bioassay or had the same amount of T1 are scored as
neutral.
...................-... ............ .. .............................................. .....
i I
Higher amount of T 1 j Lower amount of T 1 I Neutral i
1....................................... 1....
| Trials with non- | 1 10
J signallers | 9 I1
0
|
1 Trials with | 4 I\ 3 1 4 I
| signallers only 1 \
1..................................... r ■”
j All trials | 13 i 3 I 114A |I
...... «..._____ ___...__ _______i... ----------------------------,---L*.
Table 4.4.5c: Frequencies of winners of bioassay trials in terms of relative amounts of T2 detected in their
pheromone signal. Trials where the males drew in the bioassay or had the same amount of T2 are scored as
neutral.
t — — ...................
1| | Higher amount of T2 | Lower amount of T2 j Neutral j
.j.................................
f ---- \1—
....... —
i
1 Trials with non- I 9 0 1 10 (
| signallers 1 j
|
j
Trials with
signallers only
I 6
I1I 1 II I 4
■■■■ j —— —i— -----
......—•— ——- ........j— ------------------------- --\- ----
---------------------------
All trials 15 I 1 j 14
Table 4.4.5d: Frequencies of winners of bioassay trials in terms of the relative ratio of T1 and T2 detected in
their pheromone signal.
| : |
Higher ratio (T1/T2) j Lower ratio (T1/T2) ! Neutral j
Trials with | 3 | a i
4 1
signallers 1I \1 1 %
Wi:-LL -L-f-crio-cr. .oocnninof^
59
Chapter 4: Inter-male variation {I bioassay and II chemical analysis)
To summarise, the amounts of both T1 and T2 were found to be a good

predictor of the outcome of a bioassay trial, however, if those males subsequently found
not to be signalling are taken out of the analysis, not enough replicates remain to determine
which characteristics of pheromone emission are preferred.
4.5 DISCUSSION
Both males and females have been shown to detect fairly subtle differences in
an aggregation-pheromone signal and to vary their response accordingly. Neither males or
females make an absolute choice between two signals. Instead, responders bias their
response away from a random one to a degree that is related to the difference between the
signals. All or nothing responses are predicted by Eberhard (1996) for cases where
response is governed by natural selection. This hypothesis is proposed in the context of
mating cues delivered to the female in the ejaculate, but could, I think, equally be applied to
other signals between the sexes. Eberhard proposes that dose-dependent responses are a
characteristic of sexually selected cues since they allow females to choose between males
either for ‘good viability genes’ or ‘good attractiveness genes’. Males in this system also
give a dose-dependent response to other males. Perhaps choosing the same way females do
will maximise their female encounter rate.
In this series of experiments we have shown that a predictable outcome of trials

can be obtained using this bioassay method (fig. 4.4.1). Therefore, concern regarding
sources of error that could obliterate any pattern in the data arising from the method have
been allayed. We have also shown that the response of one beetle to two signals can be
used to predict the response of other beetles given the same choice. The accuracy of these
predictions is highest when the sex of the responders is kept constant and time between
trials is kept to a minimum. Observable variation in male signal remains when strain, age,
ambient conditions and culturing conditions are kept constant.
In some experiments presented in this chapter all possible combinations of male

pairings were trialled and consistency was measured by comparing the outcome of single
trials with the overall-score difference of the two males. Although this method does give
satisfactory results, perhaps a better experimental design is that employed in experiment
4.3.4. In this design consistency of result was measured by simply repeating trials. The
design used in 4.3.4 has two advantages over the earlier design. First, more individuals can
be screened for the same number of bioassay trials. Second, should any signaller die/escape
during the trials, a smaller portion of the data is lost. The calculation of overall score
requires all trials to be completed successfully, the loss of any single insect therefore,
automatically eliminates all data derived from all trials involving that insect.
60
Male response was found to be only 56% that of the average female response
in LGB (compare fig. 4.4.2a to fig. 4.4.2b). This would seem to indicate that the opposite
sex to that of the signaller is more responsive to the signal, thus supporting the sex-fiinction
hypothesis. However, in a similar study of the saw-toothed grain beetle, Oryzaephilus
surinamensis (L.), White and Chambers (1989) highlighted that, “as males produce the
pheromone, females would not have been exposed to pheromone prior to the tests, in
contrast with the males, and this could be responsible for any observed differences in their
responses”. White and Chambers (1989) controlled for this possible habituation of
responders by culturing males and females together up until the point of testing. They
found that females always showed a higher response to the male-produced pheromone
whether cultured individually or in a mixed sex culture. However, changing the culturing
environment did have a significant influence on the difference in response of the sexes since
males cultured individually gave a lower response than those in a mixed culture with the
reverse being true of females. The observation that a male’s response is increased by the
presence of the opposite sex suggests that perhaps males decrease their signal in response
to the presence of females, a possibility not mentioned by White and Chambers (1989).
Unfortunately their experimental design incorporates mating as a confounding

variable. Mixed cultures will not only allow males and females to share the same physical
space, they also allow for mating to occur prior to testing. In this study this possibility was
eliminated by keeping insects separate, yet allowing free exchange of gases through the use
of gauze pot lids. Here we have concentrated on the influence of prior exposure of
pheromone on females and have not investigated the possibility of increased male response
due to removal the signal he is making.
Borden (1967) also investigated the effect of previous exposure to pheromone

on subsequent response to pheromone in the bark beetle, Ips confusus. It was found that
female response to aggregation pheromone was significantly lowered after just 15 minutes
exposure to male frass (presumably emitting pheromone signal) compared to a clean air
treatment. Although the response of males was also lower after previous exposure, this
change was not significant. Perhaps the males used as responders in this experiment were
still signalling and therefore the clean air treatment they received became contaminated by
their own signal.
In summary, White and Chambers (studying Curculionidae) found female

response to be lowered by 9% by rearing females in a mixed as opposed to single sex
culture. Borden (studying Scolytidae) found female response to drop by 73% after just 180
minutes exposure to male aggregation pheromone (presented as male frass). LGB females
in this study gave a 51% drop in response if exposed to higher rather than lower levels of
pheromone for 8 days. This large drop in response is more than enough to explain the
earlier observation that males exhibit only 56% of the level of response that females do to
61
Chapter 4: Inter-male variation (I bioassay and // chemical analysis)
the pheromone signal (44% drop relative to the female response). Therefore, no sex
differences in perception or reaction to aggregation pheromone have been demonstrated in
this study. Scholz recorded a similar habituation of female LGB exposed to artificial
sources of T1 and T2 (Scholz 1997), however in this study, habituation was only recorded
when females were removed from maize. Perhaps starved females are even more readily
habituated than feeding females.
Bioassay trials linked to chemical collections of pheromone plumes showed

that the bioassay reliably picked out which males of the 10 used were not signalling.
Unfortunately, not enough replicates remained to give any further information about the
chemical characteristics of preferred signals.
To conclude, conspecifics are sensitive to variation in the pheromone signal.

Aggregation-pheromone characteristics could therefore easily be selected as a component of
mate choice in this species. The degree to which conspecifics actually choose between two
competing signals, in the field, is investigated in the next chapter. Male signallers can vary
their investment in signalling with time, therefore this character is very flexible. Cade
(1984) suggests that, “...facultative alternatives and underlying genetic variation are not
mutually exclusive possibilities...individuals may differ in the tendency to change
behaviour under different conditions”. This means that the total investment in aggregation-
pheromone signalling per male could be fairly complicated to calculate. Any attempt in
constructing a hierarchy of attractiveness between males is therefore only really practical
over a short time scale (in the order of a few days). This knowledge helped to design later
tests to determine some of the consequences of signalling during courtship.
62
Chapter 5: Inter-male variation (111field trials)
CHAPTER 5: INTER-MALE VARIATION IN PHEROMONE

SIGNAL (III: FIELD TRIALS)
5.1 INTRODUCTION
Synthetic aggregation pheromone is used to bait flight traps that are widely
distributed to monitor the spread of Prostephanus truncatus across Africa (see chapter 1).
Considerable inter-male variation in naturally produced aggregation pheromone signals has
been found both by chemical analysis of signals (Prof. David Hall pers. comms.), and
through the use of bioassays in Prostephanus truncatus (chapter 4 this study). It was
shown in chapter 4 that test females generally agree which males are most attractive and
males show the same preferences as females. It was not known how well these patterns of
preference found in the bioassay apparatus would translate into patterns of dispersal in a
natural situation. Also, there is always the worry that populations maintained for many
generations in the laboratory may be inbred, or exhibit characters that do not reflect those
found in natural populations. Field trials were conducted in order to address these
problems.
Single males were used as lures to bait flight traps in a woodland habitat in Ho,
Ghana. A funnel-trap design was used since this design has been shown to be relatively
effective at trapping both LGB and Teretriosoma nigrescens compared to some other
popular designs (Key et al., 1994). Traps were arranged in pairs and as such acted as mini
choice-tests. This allowed collection of data on the level of choice conspecifics make
between the pheromone signals of two males signalling side by side. Fadamiro noted that
P. truncatus is often seen to hover over a pheromone source (Fadamiro, 1995). It is
possible that this behaviour could allow dispersing beetles to evaluate aggregation
pheromone signals in some way. By sexing all beetles caught in traps, sex differences in
preferences could also be assessed. Volatile samples were collected from male signallers in
the field with a view to correlating pheromone output and component ratios with signal
success.
The numbers of the predator, Teretriosoma nigrescens, caught in traps were

also of interest. The predatory beetle Teretriosoma nigrescens, released as a biocontrol
agent against LGB, first arrived in Ho around 1996 (R.J. Hodges, pers. comm.).
Teretriosoma nigrescens follows LGB’s aggregation pheromone to locate its prey. The
predator had already been found in traps baited with artificial sources of pheromone (Tigar
et a l , 1993), but no estimates of numbers attracted to natural sources were known. Ghana
was chosen as a study site, mainly for logistical reasons. There is already an established
programme of research on Larger Grain Borer at the Ministry of Agriculture in Ho, Ghana.
63
Chapter 5: Inter-male variation (III field trials)
This field work was also timed to coincide with trials planned by Dr. Rick Hodges
travelling from the UK.
Two studies already exist where single male Prostephanus truncatus have been
used as lures in flight traps. Wekesa (1994) placed single-male baited flight traps about
60m away from a source of P. truncatus in Kenya. Significant numbers of conspecifics
were not attracted, but there were only relatively low numbers of flying insects present.
Significant trap catches were, however, obtained using single-male lures in a study
performed in 1995 in Togo (Scholz, 1997). In this study single males were allowed to
tunnel into maize cobs that were then hung inside delta traps with the glue removed, and left
for varying lengths of time from one to four weeks. After the trapping period, numbers and
the sex of beetles colonising the cobs were determined. By letting beetles accumulate on
cobs, this experiment was able to assess the changes in attractiveness of a growing
aggregation of beetles. Traps left for a week attracted a mean of 59 beetles per week
(median 39). The sex ratio of beetles attracted was female biased (64% females). The work
reported in this chapter allows an estimation of the number of conspecifics a single
signalling male can attract each day, and thus gives an idea of the efficiency of signal and
response in this species in the field.
5.2 AIMS
• Estimate the number of beetles that can be attracted to the immediate vicinity of a single
signalling male per day in a woodland site in Ghana.
• Estimate the sex ratio of beetles attracted by a signalling male.
• Estimate the level of choice being made by responders. Are some signallers more
attractive than others if they are presented side by side? If so, what is the magnitude of
their advantage in the field?
• Evaluate any sex differences in the level of choice being made by responders between
signals.
• Estimate the variation in the quantity and quality of pheromone emitted by males taken
from a Ghanaian population.
• Test for any correlation between natural variation in pheromone signal (chemically
determined) and natural variation in attractiveness of these emissions (trap data).
64

5.3.1 Trapping trials:
Study site:
An area of teak plantation situated on the side of a south facing ridge on the
outskirts of Ho, Ghana, was used as a study site. This study was performed in February
1997. Daytime temperature ranged from approximately 25-35 °C. Humidity was generally
low for the area and ranged from approximately 40-70% r.h., peaking at night.
Insects used:
Males used as signallers in these experiments were taken from two sources.
Initially all males were trapped directly from the study site using a trap baited with an
artificial pheromone lure (polythene vial containing lmg T1 and 2mg T2). Subsequent
waves of signallers used were taken from cultures of locally caught beetles reared on maize
at the Ministry of Agriculture in Ho. Beetles were taken from the top of three month old
culture jars.
Trapping equipment:
Traps: Flight traps were used to catch beetles in these trials. Initially both a
simple funnel trap and a funnel trap with a baffle were tested (see fig. 5.3.1a).
Subsequently, only the simple funnel traps were used. Where traps were presented in pairs,
they were suspended from a single point and separated using a length of stick (see fig.
5.3.1b and fig. 5.3.1c). The wires used to attach traps to trees were coated in insect glue to
prevent walking insects such as ants gaining access to the male-on-food lure.
Lures: Grains containing male signallers were placed individually in open small
glass pots, covered in fine nylon mesh (100 holes per cm2) to avoid insects getting in or
out. One extra perforation of approximate diameter 2mm was added to facilitate the
emission of any pheromone through the mesh (see fig. 5.3. Id). Lure pots could easily be
swapped between traps since they were attached using Velcro tape.
65
SIMPLE FUNNEL TRAP SIMPLE FUNNEL TRAP

WITH BAFFLE
wire dipped in
insect glue
25cm plastic
baffle
18 cm
22c:
funnel
lure pot
collecting pot
top :8cm xl0.5cm __
bottom:5.5cmx8cm water
depth: 9.5cm
Fig. 5.3.1a: Diagram to show the two trap designs that were tested: simple funnel trap and simple funnel
trap with baffle.
branch o f tree wire dipped in

insect glue
stick
60cm
sim ple funnel traps
Fig. 5.3.1b: Diagram to show how trap pairs were suspended from trees separated by 60cm using wooden
sticks.
66
Fig. 5.3.4c: Photograph o f one trap pair in the study site (Teak plantation in Ho, Ghana).
Chapter 5: Inter-male variation (ill field trials)
one hole of approximate diameter 1mm to

facilitate pheromone emission
nylon mesh to
prevent entry/
exit of insects velcro tape for easy
yet allowing attatchment to trap
1
transfer of gases
whole maize
grain frass created by
tunnelling male
glass pot (cylindrical:
height=2.5cm,
diameter= 1.3cm) single male
signaller
Fig. 5.3.Id: Lure pot containing a single signalling male feeding on a small quantity of maize.
EXPERIMENTAL DESIGN:
Preliminary trials
Initially, 21 traps were hung singly approximately 25m apart in the study site in
two parallel lines. Each trap was hung from branches of trees in the teak woodland
approximately l-2m above the ground. Each trap was randomly assigned to one of three
treatments: simple funnel trap with baffle and single male lure; simple funnel trap with
single male lure; and simple funnel trap with baffle but no lure. All males used in lures were
wild-caught males who had been previously placed in fresh grain for three days. After one
day of trapping, all traps were emptied and all Prostephanus truncatus and Teretriosoma
nigrescens were counted. All individuals of Prostephanus truncatus were sexed as
described in section 1.5.6.
All traps had to be temporarily removed from the study site away from the
threat of bush fires one day after they had been put in place. Traps were redeployed after a
delay of one day in pairs as required for the main choice trial.
Choice trial
Adding baffles to the trap design did not perceivably increase trapping
efficiency, so the simple funnel design was used for all subsequent trials. The simple
design has the advantage that beetles are less likely to be knocked into a trap while possibly
hovering around comparing pheromone signals.
68
Traps were hung in pairs with a distance of 0.6m between traps and at a height
of about 1.5m above the ground (see fig. 5.3.1b). Initially eight pairs of traps were set up
in a line parallel to the road. Pairs of traps were never closer than 25m from each other and
were generally approximately 30m apart. Both traps were baited with a single male lure in
four pairs and only one of a pair of traps was baited in the remaining four replicates.
The aim of having replicates where one trap was baited and the other left empty
was to assess whether beetles might blunder into a trap while following the lure of an
adjacent trap just 0.6m away. In order to maximize numbers of beetles attracted to these
replicates, one lure that caught no beetles was replaced with a fresh one. After four days of
trapping (5th Feb.) all eight replicates were given lures in both traps. On the 9th of Feb., all
traps bar one had to be removed as a bush fire swept through the site. The bush fire was
fueled by leaves and dry brash mainly on the woodland floor and only caused superficial
damage to most of the trees. After a one day delay, nine pairs of traps were redeployed.
These were in place for 11 consecutive days of trapping.
Traps were emptied once each day between the hours of 8am and 3pm and in
the majority of cases between 8am and 10am. Numbers of Prostephanus truncatus and
Teretriosoma nigrescens caught per trap were recorded and all Prostephanus truncatus
caught were sexed.
Lures were generally assigned to a pair of traps and remained there until
removed to be placed in the volatile-collection equipment. Each pair of lures was switched
between the two possible positions in a pair of traps each day in order to allow an
assessment of position effects within a trap pair. As trapping data and volatile samples were
taken from pairs of males, new pairs of males were introduced in their place. A total of 28
wild caught males were used as lures, of which 18 were sampled for volatiles. In addition,
21 males taken from local insect cultures were used as lures, 12 of which were sampled for
volatiles.
Analysis:
Details of the analysis are presented in the results section. Generally, chi-
squared analyses were used to test hypotheses regarding the distribution of beetles caught
among the traps. In some cases some low trap-catch data were not used in the analysis in
order to fulfill the recommendation that at least four fifths of the expected values should be
above five (Sokal and Rohlf, 1981).
5.3.2 Pheromone samples:

Collection:
Collection filters were made as described in section 4.3.5.
69
Pheromone samples were collected using the apparatus shown in fig. 5.3.2.
Note that volatiles from each of a pair of males were collected simultaneously. Each sample
consisted of 23-25 hours of continuous collection. Air flow through each filter was
approximately 1 litre per minute, however electricity supply to the pump did fluctuate
enough to affect its performance and there were a few brief power cuts. The activated
charcoal was replaced twice throughout the collection period. Once collected, volatile
samples held on filters were stored in one of two freezers at approximately -20°C. Two
pairs of control samples with no lure added to the collection chamber were taken. Samples
were then transported to the Natural Resources Institute, Chatham, Kent, UK for analysis.
Chemical analysis:
The volatiles collected were analysed as described in section 4.3.5.
70
air chamber (to

smooth air flow )
poropak
filter
plastic pheromone
collection chamber
volatile source pot

plastic with mesh removed
tubing
cylinder o f activated
charcoll (to clean
incoming air)
3cm
I
air flow (approx. 3 litres per minute)
Fig. 5.3.2: Plan view of the volatile collection apparatus.
5 .4 RESULTS
5.4.1 Trapping trials
Preliminary trials
Prostephanus truncatus were caught in 10 o f the 13 baited traps. N o P.
truncatus were caught in any o f the unbaited traps. Mean catch per trap and the associated
standard errors for all treatments are shown in fig. 5.4.1a. Sim ple funnel traps caught
comparable numbers o f beetles as traps with a baffle.
71
to N=7
a
ig
Sx
3a N- 6
-s: &
—4 -
§•&
r
a. a
*- 3 i
3
o JS
i-
<u 6 30
JO Cd
B
3C «
3cd
<D N=8
Trap with Trap with Simple trap

baffle and baffle but no with lure
lure lure
Fig. 5.4. la: Mean number of Prostephanus truncatus caught per trap per day for each of three trap types:
funnel trap with baffle baited with a single male; unbaited funnel trap with baffle; simple funnel trap baited
with a single male.
Choice trials
N o P. truncatus were caught in unbaited traps that were paired with baited
traps. Baited traps caught an average o f 4.7 beetles per trap per day (N =16 : four traps for
four days).
The frequency distribution o f numbers o f LGB caught per trap per day (for
traps that were part o f a trap pair) is skewed to the left (m ean=4.15; m edian=2, 25%
quartile=0, 75% quartile = 6, N =275, all LGB per trap per day). The maximum number o f
LGB caught per trap per day was 31. The proportion o f the 1150 P. truncatus caught in
these traps that were fem ales was 0.64 and daily estimates ranged from 0.59 to 0.73.
Average daily trap catches and their associated sex ratios are shown in fig.5.4.1b. Beetle
numbers were not randomly distributed between the two traps o f a trap-pair. A nx2 chi
square analysis with Yates correction combining data from all trap-pair catches that caught
at least tw o beetles gave a sum o f chi-squared o f 207 with d.f.=90 and therefore p<0.001.
72
Chapter 5: Inter-male variation (III field trials ')
4 0.8
■hao 2 0.7
uCu £’ea
E
a. 0. 6 £
0
2CO
8L 0.5 «
OQ 8
a c
0
j 0.4
0 n
6
1
E
0.3 1C-
33 4 _o
'3
<D 0.2 2
2o x
u
on
> 2
<
0 0
CM CVJ CM W CM CM CM CM CM CM CM CM CM CM CM CM CM CM CM
CM CO m - in CD 00 o> CM CO ^ If ) CD ha CO O) O T-
T~ T~ T~ -- T- CM CM
Date
cz=3 Average number of LGB per trap per day

♦ -Sex ratio (proportion that are female)
Fig. 5.4.1b: Chart to show the average number of Prostephanus truncatus caught per trap per day and their
sex ratio, for traps set in pairs. The number of traps sampled is indicated as a number above the error bar.
G iven that one trap in a pair catches more beetles than the other, how much of
this skew is associated with the signaller? Data where total trap catch for a pair o f traps was
at least eight individuals for two consecutive days, and where the pair o f signallers was
swapped over between the two trap locations at the end o f the first day were used for this
analysis. 27 sets o f data fulfilled these criteria. The data comprise results from 11 different
pairs o f males and nine trapping sites. The data were tabulated as a series o f 2x2 tables,
(see table. 5.4.1c).
Table. 5.4.1c: Example of 2x2 table used to distinguish day, trap and male effects on trap catch distribution
between two traps of a pair.
TRAPA TRAPB
DAY 1 Total caught by MALE 1 Total caught by M ALE 2

I .........uirrr- f
,........... ~ n ........^ .......... , | ________ _____ r ....... -..... ~ - t ................. — -------r -T........ ................ -................................................
DA Y 2 Total caught by MALE 2 Total caught by M ALE 1
For each 2x2 table o f data three nx2 chi-squared statistics with Yates correction
were calculated to distinguish between the three possible effects detectable using this table:
73
1) Day effect
2) Trap effect
3) Male effect
It was found that day, trap position and male identity all significantly affected
trap catch (see table 5.4. Id). Male identity was found to the most influential factor.
\ 1
1 | Zchi-squared | degrees of freedom | probability |
i i 1
T.......—... —1
j Day effect | 44.8 | 27 J 0.02 1
| Trap effect 1 67.9 | 27 1 « 0 .0 0 1 I
i....... .....* ’
| Male effect | 87.5
o
o
o
i—
| 27
v 1
H
.... ....................... .... i................ ....................
]
Table 5.4. Id: Summary of determinants of skew between traps of a pair.
A 2x2 contingency table with Yates correction shown below (Table 5.4. le),
was used to ask the question, do females choose between traps in a pair differently from
males.
j Male Signaller 1 | Male Signaller 2 J

■i ■■
'...................... —..|
| No. of Males j a | b I
j No. of Females | c I d I
...-....j
Table 5.4. le: Example of 2x2 contingency table used to test if males and females choose between two traps
of a pair differently.
Combining all cases where the trap catch per day was at least nine males and
nine females and at least one beetle was caught by each trap gives nine replicates. The sum
of chi-squared for the interaction between sex and trap catch is 2.4 (d.f.=9 and p>0.95),
which shows that no significant difference between the choice of males and females has
been demonstrated.
Differences in the choice of males and females can also be studied on the next
spatial scale up, between pairs of traps. Since there is variation between total catch per pair
of traps it would be interesting to see if the sex ratio of beetles caught varies with changing
size of trap catch per pair. A greater proportion of females in high trap catches would
indicate the possibility that females show a higher level of choice than males. Data collected
from the 6th Feb. to the 20th Feb. were used since trapping during this time was relatively
consistent between days. Trap catches were ranked in terms of the numbers of beetles
caught. The tally of beetles was then divided up into eight consecutive sections of about
110 beetles. The sex ratio and average trap catch per pair of traps were calculated for each
74
section. Using the null hypothesis that sex ratio does not change with size of trap catch,
expected numbers of females caught in each section were calculated. A chi squared test with
8 d.f. gave a sum of chi-squares of 5.0, p=0.76, which shows that no significant variation
of sex ratio could be detected with changing trap-catch size.
5.4.2 Pheromone samples

Unfortunately collection of volatiles was unsuccessful. T1 was difficult to
quantify due to large amounts of impurities in the samples. The amount of T2 was
quantified and ranged from 0.005-0.09|ig per hour, but control samples contained
comparable amounts of T2 to test samples: Test sample average = 0.033jig per hour;
control sample average = 0.03 l|ig per hour (with no male signaller in the collecting
chamber). Therefore these samples could not be used to evaluate chemical variation in
pheromone signal.
5.4.3 Additional observations

An average of 0.5 Teretriosoma nigrescens were caught per baited trap per
day. Trap catch was very unevenly distributed between days with approximately one third
of the total caught on a single day. No Teretriosoma nigrescens were caught in unbaited
traps.
Approximately 20 hours were spent searching for Prostephanus truncatus

tunnelling in vegetation in the study site. Although other Bostrichid species were located,
no P. truncatus were ever found. Test tunnels were however, found in the sticks used in
the construction of trap pairs. A total of seven tunnels were found, five of which were into
the cut end of the stick. Tunnels were found in both types of wood used for trap
construction, which were shoots of teak and another unidentified tree species. Two live P.
truncatus were removed from these tunnels, all others were empty.
5.5 DISCUSSION
Considerable variation exists in trap catch both between traps and between
days. Mean trap catch was approximately four beetles per trap per day. This shows that
males are able to attract significant numbers of beetles even when signalling alone in
Ghana. It is thought that there is a sizable population of beetles away from maize utilizing
an unidentified wood host or more likely, hosts (see chapter 1). Males signalling whilst
feeding on such hosts may be less effective signallers than the test beetles in this study (see
effect of host on signal in chapter 3). Here, the males used were feeding on maize, which is
likely to be a relatively nutritious food source. The success of single-male trapping first
demonstrated by Scholz (1997) and confirmed here opens up the possibility of a very
profitable area of research into the actual patterns of signalling and response in the field.
75
Chapter 5: Inter-male variation (IU field trials)
It is apparent, however, that knowledge of the spatial distribution of this pest

away from storage will become more and more crucial to any interpretation of the
population dynamics of this species. It is still not known whether Prostephanus truncatus
is widely scattered or forms large aggregations in the bush akin to those occurring in maize
stores.
The overall sex ratio of beetles attracted in this study was 64% females. This is
exactly the same sex ratio reported by Scholz (1997), attracted by male-baited lures in
Togo. A female bias has also been found in flight traps baited with artificial pheromone
lures (Hodges et al., 1998; Scholz et aL, 1997a). This implies that females predominate in
the dispersing population, and/or that dispersing females are more likely to follow
aggregation-pheromone plumes than dispersing males. Laboratory studies have shown that
females show a higher response to naturally produced pheromone in a walking bioassay
(Hodges and Dobson (in press); chapter four this study), but this may be due to male
habituation to pheromone which would be less likely to influence dispersing beetles.
Fadamiro (1995) reported no differences in flight activity or duration of flights between the
sexes. This result is contradicted by the female-biased flight activity recorded by Scholz
(1997). Fadamiro also found that there was no sex specific differences in response to
artificial pheromone sources in wind tunnel trials when insects had already been
preselected for flight activity. Li (1988) found that the sex ratio at emergence was highly
significantly female biased (58.5% females), however Scholz reported a 1:1 sex ratio in
insects cultures (Scholz, 1997). It is therefore unclear whether females predominate in
samples attracted by pheromone simply because they are more abundant in the overall
population, or because they show any sex-specific patterns of dispersal or response to the
signal.
The female bias in the dispersing population may be dependent on habitat type.
Scholz highlights that sex ratios in wooded habitats away from stores are often of almost
equal sex ratio (Ramirez-Martinez et al., 1994 quoted in Scholz, 1997). She proposes that,
“the higher nutritional value of stored food products may act as an arrestant, making males
(for reasons unknown) less likely to react to other cues... that would otherwise have
stimulated them to migrate” (Scholz, 1997).
Males could increase their reproductive success by employing one of two

strategies. They can signal for mates by emitting aggregation pheromone or they can follow
other males’ signals. In this study males caught using an artificial aggregation-pheromone
signal were subsequently successfully used to provide a pheromone signal. This shows that
males cannot be divided up into two mutually exclusive categories of signallers OR
responders, but may utilise both strategies.
76
The female bias in trap catch supports the idea that aggregation pheromone in
Prostephanus truncatus may be maintained through an increased mating advantage for
signalling males. A highly significant skew in trap catch between two traps of a pair was
found. Identity of the signaller was found to be the most significant determinant of trap
catch. The large influence of male identity on trap catch implies two things: that variation
exists between male signals; and that this influences the response of conspecifics. This
result agrees with equivalent bioassay trials in the laboratory (Chapter 4 this study). In
bioassay trials, males and females were found to be equally discriminating between signals
from two competing males. The field trial carried out here has given the same result. Males
and females therefore seem to invest equally in the assessment of signals and there may be
no special sex-specific adaptations to choose between aggregation-pheromone signals.
One sex-specific difference in choice between signals does exist: this is the
significant increase in female bias in trap catches in traps baited with artificially produced
T1 and T2 (lmg of each) compared to traps baited with each component alone (lmg total)
(Scholz, 1997). Scholz proposes that this could arise if males land at a lower threshold of
response to concentration of pheromone. This could be the case if males follow pheromone
primarily to locate new hosts and females locate the point of highest concentration to locate
the signalling male as well as a host. Indeed males may actively avoid landing exactly
where another male competitor has tunnelled if the chances of him silently gaining access to
mates is low (see the discussion in chapter 6). It is difficult, however, to extrapolate from
results obtained using synthetic pheromones at abnormally high concentrations to the
natural situation.
Collection of pheromone was required to quantify chemically inter-male

variation in aggregation-pheromone signal. T1 peaks were masked by large peaks of
impurities. During the trials, many bush fires were lit in the area which may have swamped
the samples with impurities. The highest peaks of impurities masking the T1 peak were
collected on the same day as mbbish was burnt about 20m away from the volatile-collection
apparatus. A similar amount of activated charcoal was used to clean incoming air as that
used by Prof.D. Hall at NRI, but this may have been insufficient for the conditions found
in Ghana.
T2 was quantified, but samples contained comparable amounts to those found

in the control samples (no male present). The equipment designed for collection of
pheromone in the field may have been at fault. Although the basic design was copied from
apparatus used by Prof.D. Hall (described in section 4.3.5), which has been successful,
some glass parts were exchanged for more robust plastic equivalents that may have retained
some of the pheromone and/or created too much turbulence in the airflow, leading to
inefficient absorption of pheromone onto the filter. Lastly, the ability of poropak filters to
trap volatiles decreases with temperature. Although samples were kept refrigerated in
77
freezers (-20°C) whilst in Ho, they were exposed to much higher temperatures during
transportation back to the UK, especially during the journey to Accra.
Aims one to four were successfully completed. This shows that field studies of
natural pheromone plumes are relatively easy. Any future studies in this area can also be
expected to be fruitful. Collection of volatiles from single signalling males was
unsuccessful. Possible remedies for this include: an increase in the efficiency of cleaning
incoming air; avoidance of the bush fire burning season; improvements in the design of the
collection chamber; and refrigeration of samples during transport back to the UK
78
Chapter 6: Mate choice on contact
CHAPTER 6: MATE CHOICE ON CONTACT
6.1 INTRODUCTION
Work has already been presented in previous chapters that describes some of
the factors influencing how beetles will come in contact with each other. This chapter
presents work on how they then interact, and the role of aggregation pheromone in these
encounters. In other words does aggregation pheromone in LGB also act as a courtship
pheromone? The focus is on behaviours that determine differences in mating success
between individuals. In any aggregation of reproductively active conspecifics, both
interactions within the sexes and between the sexes are possible factors leading to mating
preferences. Since this thesis is concerned with differences in males (more specifically their
aggregation signal), female choice and male contests were studied. Female choice is used in
this study to mean any process, be it involving her nervous system or not, that leads to
female-determined selection between potential mates (see chapter 1 of Eberhard, 1996).
Male contests are also defined in general terms and need not be mediated through physical
contact. Sexual selection between females (male choice and female-female contests) has not
been specifically investigated in this study.
Long-range pheromones produced by males that are attractive to females have

been found to influence contact courtship behaviour in the cockroach, Nauphoeta cinerea
(Moore et aL, 1995 and 1997, and Moore, 1988). Manipulation of the pheromone signal
has shown that dominance amongst males is determined by characteristics of this
pheromone (Moore et al., 1997). Female preferences of male odours also correlate with
those that determine the male hierarchy (Moore, 1988). Moore’s studies show that a long-
range olfactory signal can continue to be used as a social cue at close range. Close-range
pheromone cues have already been documented as sexually-selected characters in
Lepidoptera (refs, in Andersson, 1994).
Since male Prostephanus truncatus have the ability to detect aggregation

pheromone signals, it is possible that the pheromone plays a part in male-male interactions.
Both visual and acoustic signals have frequently been shown to be instrumental in
interactions between members of the same sex in other species (see chapters 13 and 14 of
Andersson, 1994). Acoustic signalling in Anurans (frogs and toads) and Orthopterans
(crickets and grasshoppers) has been shown to influence the spacing of males (see refs,
quoted in Andersson, 1994).
Little was known about the courtship behaviour of Prostephanus truncatus. Li

(1988) never observed Prostephanus truncatus mating and concluded that this must occur
out of view, in tunnels. Fadamiro (1995) however, observed mating in an open arena with
79
no food present. Fadamiro reported that, “the male was the mate-finding sex: females did
not attempt to locate mates”. LGB courtship both inside tunnels in a potential host and away
from the host are described in this thesis. Quantification of variables associated with mate
choice has been limited to the situation where adults are away from the plant host. The open
arena provides a uniform platform where two males can easily be presented to the female at
once, and where starting position is unlikely to be very influential on the outcome of the
mating trial. Male position and behaviour within the tunnel system in a host is an area that
warrants its own study, but incorporates unwelcome confounding variables to an
investigation into the effects of pheromone signal and male size on courtship performance.
Time-lapse photography and artificial hosts sandwiched between plates of glass (for easier
observation) were used to describe adult behaviour in tunnel systems in the host.
Male size is a common determinant of mating success in arthropods (Crespi

1989). Both advantages for larger males and smaller males have been documented (see
Andersson 1994). Smaller males may sometimes be more manoeuvrable and agile in
obtaining mates, particularly if copulation takes place in the air or in other three dimensional
media. Larger males may perform better where contests of strength, either between males
or between the male and female, are involved. Larger males may also be selected visually.
Females may benefit by selecting large males if this leads their female offspring to be larger
as larger females are often more fecund (see Holloway and Smith 1987). Male size was
often controlled in the experiments presented in this chapter such that each of a pair of
potential suitors presented to a female were of different sizes. This has a two fold
advantage: first, it increases the likelihood that there will be variation in the aggregation
pheromone signal between the pair of males; second, it ensures there is variation in size
between the pair of males.
When adult Prostephanus truncatus encounter each other in an open mating

arena they will often push each other (pers. obs.). This behaviour was particularly
prevalent when females came in contact with males (see phase 1, section 6.3). Pushing is
easy to observe, is widespread, variable, and demands explanation since it appears to be
energetic and time consuming. For these reasons pushing behaviour was investigated
further. Data describing which beetles initiate physical encounters and the outcome of such
encounters helped unravel the motivations that result in mating preferences. Behaviours
occurring early in courtship have the advantage that data are more often available for both
males of a pair of possible suitors, males can thus be compared directly, and variation in
female behaviour can then be eliminated from the analysis.
Pushing behaviour between males and females could enable beetles to assess
each other’s quality, prevent copulation behaviour, and/or serve to stimulate sexual
behaviour. Pushing between members of the same sex may be a form of intra-sexual
selection, performed either to increase direct access to mates or to defend resources like
80
tunnel systems. Courtship behaviours are not necessarily all sexually selected traits, as
Andersson (1994) points out, “many aspects of courtship...may reduce escape responses
...synchronise endocrine reproductive functions, or coordinate the behaviour of mates in
space and time for copulation”.
This chapter starts with a verbal description of Prostephanus truncatus

courtship behaviour observed in open mating arenas and in tunnel systems within an
artificial plant host. Preliminary mating trials in an open arena are then presented. The aim
of these trials was to test whether measures of courtship behaviour are characteristic of
different individuals. The relative influence of the two sexes was also investigated i.e. does
female identity determine most of the variation in courtship behaviour, or does male
identity? A certain level of predictability of courtship measures is required to justify any
attempt to relate these measures to characteristics of pheromone signalling. The next two
experiments were designed to unravel any interaction between courtship success (mediated
through female choice and/or male-male interactions), male size, male pheromone signal,
and male identity. In these experiments two possible male competitors are presented to a
female. Hypotheses were tested using two main types of data in this chapter: ultimate
choice of male mate, and behavioural measures during courtship. The basic questions, ‘do
females preferentially mate with some males rather than others during courtship?’ and if so,
‘is this influenced by male aggregation pheromone emission?’, can be most directly
answered by looking at ultimate mate choice between males. Measures of behaviour were
then used to suggest possible mechanisms for patterns of mate choice.
6.2 AIMS
• Describe behaviour between adult P. truncatus within the plant host and in an open
arena.
• Evaluate whether behaviour during courtship leads to non-random mating in

Prostephanus truncatus.
• Test whether male’s success during courtship correlates with his performance in the
pheromone bioassay.
• Test whether manipulation of the male pheromone signal influences courtship

behaviour.
• Describe male-male interactions during mating trials where two males are presented to
one female.
81
6.3 DESCRIPTION OF COURTSHIP BEHAVIOUR IN P.

TRUNCATUS WHEN OUTSIDE THE PLANT HOST.
All the following observations were taken from beetles placed in a mating arena
with a flat surface with no food available (a petri dish, diameter 9.5cm, with its floor lined
with filter paper).
PHASE 1
Both the male and female walk around the arena. Encounters often appear to
happen by chance, others result from directed movement of both sexes towards each other.
Females appear to be more aggressive (in terms of physical pushing) than males at this
stage. Females often accelerate towards males and push them with their prothorax/head. If a
male approaches a female then this is most likely to result in female aggression if the male
approaches her head to head. If a male approaches from behind then this can result in
female aggression or phase 2.
PHASE 2
The male approaches the female from behind and starts to antennate (flap his
antennae) against the base of her elytra. Antennation often occurs in bursts of a few
seconds at a time with about one second break between bursts. This typically continues for
about half a minute. The female often starts to walk off, at which point the male may try a
new round of antennae flapping or go back to phase 1. If the female remains relatively
stationary the male may proceed to phase 3.
PHASE 3
The male then starts to climb onto the female’s back. He starts to vibrate his
legs as well as his antennae on the female elytra. He then manoeuvres himself so that his
genitalia are placed opposite that of the female. At this point the male occasionally loses
balance on the female’s back and can end up wrapped around her side or head. At any point
during this phase the female can start to walk away, this is especially likely if the male has
lost his position. If the male remains in position this may lead to phase 4.
PHASE 4
The male will continue to vibrate his front two pairs of legs and antennae for
anything from a few seconds to a few minutes. Then the male and female become very still
for anything from a few seconds to a maximum of about half a minute but most usually
around 10 seconds, I propose that it is during this time that sperm is delivered, although
this remains to be proven. Often during this time the male will sweep his antennae back to
lie against his prothorax.
82
PHASE 5
After phase 4 the female will usually start to walk off forcing the male to
dismount. Sometimes the male will dismount before the female starts to move. Occasionally
the female may become aggressive towards the male, but generally they separate and both
walk off.

6.4.1 Experiment to test if Prostephanus truncatus is capable of
mating within its plant host.
Males taken straight from culture were placed singly in pre-bored whole maize
grains and left to tunnel for one week. At the same time, adult females were placed singly in
glass pots on maize meal and also left for one week. Grains containing tunnelling males
were then placed singly in the centre of a mating arena (a petri dish, diameter 9.5cm, with
its floor lined with filter paper) and one of the females was introduced at the edge of the
arena. The arena was observed continually for 45 minutes after which time the beetles were
removed from the grain and the female’s reproductive organs were dissected and searched
for fresh spermatophores. Eight replicates were performed. In four replicates an additional
male was introduced into the arena to see if the resident male would expel this possible
competitor.
6.4.2 Observations of Prostephanus truncatus behaviour in tunnels

within an artificial host (recorded using time-lapse photography).
Construction o f artificial host:
An artificial host medium was created from a recipe adapted from that used by
Marange, Floyd and Hodges (submitted). Wheat was milled using a hammer mill with a
sieve size of 1mm. Fine flour was then obtained by sieving the milled wheat through a
4.25(im brass sieve (Endecotts). Maize flour was obtained by sieving medium ground
maize meal through a 4.25jim brass sieve (Endecotts). These maize and wheat flours were
then mixed in a ratio of 14 parts maize: 2 parts wheat: 3 parts water by volume to produce a
firm dough. The dough was sandwiched between two glass microscope slides (2.5cm x
7.5cm) which were compressed until the dough was approximately 1.5mm thick. The
sandwich was then baked in an oven at 90 °C for two hours. Slides were then left to
equilibrate to the temperature and humidity of the CTH room for two days before beetles
were introduced. Each group of beetles observed were released in a square plastic-container
(10.5cm x 10.5cm x 2cm) containing two artificial host sandwiches placed side by side.
Video camera set up:

Camera and video equipment described in section 1.5.7 were used to record
beetle behaviour in the artificial hosts. A time-lapse video recorder was used to convert 12
83
hours of real time observation into nine minutes of video tape played at normal speed. In
this preliminary investigation, behaviour was only recorded during the 12 hour light phases
of the CTH room although far red lighting could be used in the future to observe behaviour
during the dark phase.
Observations recorded:
Initially two male and two female beetles were observed for five days, then this
was repeated. Eight females and two males were then observed for six days. Lastly eight
males and two females were observed for six days.
6.4.3 Method for recording pushing behaviour

Four aspects of pushing behaviour were recorded: first, the frequency of
pushes; second the sequence in which pushes occurred; third, the identity of the pushing
beetles: and last, the identity of the beetle who initiated the encounter (who approached
whom). The identity of the beetle who initiated the encounter was of interest since it could
indicate active pursuit of mates which, in the case of females, might be influenced by
pheromone characteristics of the males. Note that strength of push was not quantified. All
this information was recorded as a series of vertical lines as shown in the annotated diagram
below (fig. 6.4.3a).
6.4.4 Preliminary investigation of courtship behaviour.

Source o f insects:
Six males and five females were isolated as adults randomly selected from a
culture. All beetles were separated into glass pots containing food and left in the CTH
room. Beetles were isolated for three and a half weeks to maximise the likelihood that they
would be ready to mate.
Apparatus:
An arena was used for all mating trials, which consisted of a petri dish
(diameter 9.5cm) with its floor lined with filter paper. A video camera was used to film the
behaviour from above (see section 1.5.7 for details of camera and video equipment). All
tests were performed in the CTH room.
All possible combinations of male-female pairs were tested. Trials were
conducted over four days. In each trial the beetles were placed into the area at the same time
and observed for up to 30 minutes or until 5 minutes after the first copulation. Filter papers
were removed after each test and the arena was cleaned with IMS. Beetles always had at
least 1 hour and 45 minutes between tests and for the majority of cases this time was
longer. Beetles were tested for a maximum of two times per day. Within these constraints,
the order of trials was randomised (picked out of a hat).
84
Order of occurrence
All encounters
identity of
male
FM
I between male 16
and the female
T All encounters
between male 32
and the female
All encounters
i between male 32
and male 16
Key to symbols
Male/female encounters: Male/male encounters:
= Female approached but no push = Large male approached but no push
^ = Male aproached but no push ^ = Small male approached but no push
F
= Male approached and female pushed I = Large male approached and pushed
L
£ = Female approached and female pushed ^ = Small male approached and pushed
= Male approached and male pushed
= Female approached and male pushed

M
= Male approached and copulated (phase 4)
T Fig. 6.4.3a: Annotated diagram of method of recording pushing behaviour between one female and two
males (numbered 16 and 32) in a mating arena.
6.4.4 Experimental design cont:

Test subjects were identified by making a note of any distinguishing features of
the beetle and the relative size of the subjects in the trial. The initial position of each beetle
in the arena was also noted to allow easy identification of the beetle on film. Video tapes
were analysed and each push was recorded in the sequence that it occurred and any
copulation attempts were also recorded as described in section 6.4.3. Potential matings
85
were inferred when males mounted females and their abdomens touched in the region of
their genitalia. Time elapsed from the start of the trial to the start of the first copulation was
noted.
Analysis:
A GLM model was used to test for consistencies in behaviour of both male and
female individuals (Minitab).
6.4.5 Investigation into the relationship between pheromone

attractiveness and success in mating trials.
Source o f insects:
All insects were randomly selected from culture as adults, sexed, and placed
singly on fresh medium ground maize in a glass pot. Insects were then left in the CTH
room for 12 days before the first assay. All males were weighed five days after being
sexed.
Apparatus:
The olfactometer apparatus described in section 3.2 and mating arena described
in experiment 6.4.4 were used.
Insects were grouped into replicates of two males and a female. Males were
paired up such that all pairings had a standard size difference of 0.4- 0.5mg. This allowed
us to investigate how any preferences between males of a standard difference in weights
varied with the average weight of the pair i.e. are larger males always preferred or are
smaller males preferred when both test insects are relatively large? The smell of the two
males was presented to the female in the olfactometer, then all three beetles were given the
opportunity to interact in a mating arena. All 36 replicates were assayed over five days.
Pheromone bioassays: Bioassays were performed as described in section 3.2.

Replicates were bioassayed in groups of three replicates. Each male of a pair of males was
connected up into the olfactometer in a position diametrically opposite the other. Position in
the arena was randomised (coin flip). Each bioassay was continued for 20 minutes.
Mating trials: Mating trials were conducted for each replicate straight after
bioassay trials (the three replicates bioassayed together were observed for mating behaviour
in sequence). Courtship behaviour was observed for 30 minutes or until one male copulated
with the female (phase 4). Behaviour was recorded as described in section 6.4.3.
This experiment was then repeated one month later using beetles from a
different, but similar culture. 45 replicates were performed in the second set of trials. This
increased the total number of replicates to 81.
86
Analysis:
For each replicate a volatile score (number of visits to source tube per assay), a
behaviour sequence record until the first mating, and the weights of each male were
recorded. From this information questions were asked about the relationship between
volatile preference and mating preference; volatile preference and weight of male; and
mating preference and weight of male.
The frequency of the six recorded behaviour types occurring between the
female and each of the two males was used to assess which aspects of pushing behaviour
between males and females are associated with mating success. Males were then
categorised as the larger or smaller male of the pair, and behaviour frequencies between
these two groups were compared. All frequency data were compared using chi-squared
tests.
Since male pairs were taken from the range of weights found in a culture
sample, it was also possible to investigate how variables changed with changing average
weight of a pair of males. Total bioassay scores for pheromone activity and behaviour
measures of courtship were plotted against absolute weight of the pair of males.
Behaviours recorded during male-male interactions were also analysed.

Differences in behaviour between the larger and smaller of a pair of males and the winner
and loser in a mating trial were assessed. All differences between categories of males were
tested using a Wilcoxon signed-ranks test.
6.4.6 The influence of manipulating pheromone signal on courtship

behaviour.
Source o f insects
All insects were those used in the second wave of experiment 6.4.3. One of
each pair of males was randomly allocated to one of two treatments:
• One fresh split grain and a small quantity of flour.
• An equivalent amount of grain previously infested with males and females

presumably containing Female Factor. This grain was taken from a two and
a half month old culture jar that was frozen for five days and then allowed
to defrost for two days.
All males were left in their treatment pots for 12 days prior to the first mating
trial.
Apparatus
The olfactometer apparatus described in section 3.2 and mating arenas
described in experiment 6.4.4 were used.
87
Experimental design
All trials were conducted over three days. The first and last days were used for
mating trials and volatile bioassays were performed on the second day.
Mating trials: Females selected randomly were placed in a mating arena with a
pair of males. Male pairings were the same as those used in experiment 6.4.5. Courtship
behaviour was recorded as described in section 6.4.3 for up to 30 mins or until one male
secured a mating.
Pheromone bioassays: Fifteen pairs of males not used in mating trials were
randomly selected to take part in bioassay trials. Both males of a pair were connected up to
the arena at once providing responding females with a direct choice between them. Female
response was recorded for 20 minutes per trial (see section 3.2 for bioassay procedure).
These pheromone bioassays were carried out to check that the manipulation of pheromone
signal had been successful.
Analysis:
Wilcoxon signed-ranks tests were used directly to compare measures of
behaviour after treatment with equivalent trials before the males were allocated to
treatments.
6.5 RESULTS
6.5.1 Experiment to test if Prostephanus truncatus is capable of
mating within its plant host.
In all eight trials females entered the grain containing the male within 12
minutes of being introduced into the arena. Generally, once entering, neither the male nor
the female left the grain before the end of the observation period. In one trial however, the
female did briefly leave and then re-enter the grain. In no case did both insects leave the
grain at the same time, therefore no matings occurred outside the grain during these
observations.
Dissection of the females revealed that six out of the eight replicates had mated
within the observation time since fresh spermatophores were found. This proves that
Prostephanus truncatus will mate within the plant host. Two intact spermatophores were
found in one of the replicates where an additional male was allowed to enter the grain. In
the other three cases where a second male was allowed entry, only one spermatophore was
found in the female. In no case were males expelled from the grain.
6.5.2 Observations of Prostephanus truncatus behaviour in tunnels

within an artificial host (recorded using time-lapse photography).
The artificial plant hosts used here were successfully colonised by test beetles.
Tunnel systems were constructed, eggs laid and larvae developed within this medium. The
88
glass/food medium sandwich was also successful in allowing observation of tunnelling

behaviours. The width of tunnels was approximately the same as the width of the
glass/food sandwich, so beetle position could be seen clearly through the glass. Details of
behaviour such as antennation and direct confirmation of penetration during copulation
were not possible using this method. The time lapse setting chosen was fast enough to
allow an easy appreciation of the activities of the beetles yet slow enough to record
copulations.
Overall distribution o f beetles within the plant host:

Generally beetles of the same sex constructed tunnel systems separately and
two beetles were only observed in the same tunnel system as a male/female pair. In the
female biased set up (8 females, 2 males), the two males still co-habited with only one other
female. Occasionally one tunnel system broke through to another resulting in more than one
male or female occupying the same system.
Some roving of beetles over the host surface (around the edge of the sandwich)
was seen. Roving beetles did enter other beetle’s tunnel systems, but generally only
remained when the system was occupied by one member of the opposite sex. Once males
had constructed a short length of tunnel (approx. 1-2cm) they spent the majority of time
positioned at the entrance to that tunnel. Lone females generally spent more time at the head
of their tunnel system throughout the whole observation period.
Pushing behaviour:
Some pushing behaviour was observed between beetles in tunnel systems and
at least one male was expelled by a female using this method. Beetles did not appear to
push each other within the plant host as much as they do in an open arena. This is difficult
to confirm, however, as beetles are more unsteady on a flat surface and therefore pushing
may be more likely to result in a greater displacement of beetles.
Male/female interactions including copulation behaviour:

Male-female pairs were observed to remain together for at least one 12 hour
light period. It was impossible to be certain whether male/female pairs were the same
individuals between days in these trials since beetles were not marked and filming was not
performed during the dark phase. Males generally remained in the tunnel entrance whilst the
female constructed blind-ending tunnels into which she laid eggs. Males occasionally
returned to the female without copulating. Beetle pairs were also observed copulating
within the plant host. In all set ups bar the male biased one (8 males and 2 females), pairs
of beetles typically copulated approximately 1-3 times per 12 hour light period. In the male
biased set up, male/female pairs were observed to copulate approximately 20 times per day.
At least two different male/female pairs copulated at these high rates.
89
The typical behavioural sequence resulting in copulations is shown in fig.

6.5.2. M ales approached fem ales and enticed them to back up into a w idened region o f the
tunnel system such as the junction between two tunnels or back to the tunnel entrance,
where there was enough space to assume the mating position. After mating the fem ale
generally returned to the end o f the tunnel, presumably to continue laying eggs and the male
generally returned to the tunnel entrance.
1. PLANT HOST
OUTSIDE
2.
3.
Fig. 6.5.2: The behavioural sequence of a typical copulation between a pair of LGB within a tunnel system
in an artificial plant host. The male is shaded in grey and the female is represented as the white beetle. Note
how the male remains in the entrance to the tunnel system before and after copulation.
6.5.4 Preliminary investigation of courtship behaviour.

G eneral observations:
Matings occurred in 22 out o f 30 trials. O f those replicates that mated, the
median tim e until mating was 8.4 minutes (average=10.1 minutes).
Both males and fem ales pushed each other using their prothorax. The pushing
beetle kept its prothorax low and accelerated against the beetle it was pushing. Pushes
90
varied in magnitude from a small nudge to the extent where a beetle repeatedly rolled
another around the mating arena. In this sample of individuals, females pushed males 10.5
times (mean), 6 times (median) before mating. On average, males pushed females just 1.2
times (mean) 1 time (median) prior to mating. This difference between males and females
was highly significant: Wilcoxon signed-ranks test, z=-4.00, N=30, p<0.001.
Time before mating and the number of encounters before mating were chosen
as two measures of readiness of the pair to mate. GLM models were used to determine if
these two variables were dependent on which individuals were used in the trial. Previous
experience of individuals was incorporated into the model. Previous experience was
defined as the number of trials already performed.
Time before mating:

The data for time before mating or until the end of the trial were not normally
distributed. The square root of all values was used to normalise the data. Initially the GLM
was run with two covariates: number of trials previously performed by the female and
number of trials previously performed by the male. These were found to be insignificant
determinants of time before mating and were taken out of the final analysis. The model used
is given below:
Vtime before mating = female + male
Time before mating was found to be more a feature of which female was used
than which male was used: Female, F=4.6, d.f.=4, p<0.01; male, F=2.04, d.f.=5,
p=0.12.
Number o f encounters before mating:

A similar model to that described above was applied to the dependent variable,
square root of number of encounters before mating, or until the end of the trial. Again, the
previous experience of males and females were not found to influence the dependent
variable and were taken out of the model. The model used was:
Vnumber of encounters before mating = female + male
Again, female identity was a more significant determinant of the measure of

readiness to mate: Female, F=7.17, d.f.=4, p=0.001; male, F=3.43, d.f.=5, p=0.021.
6.5.5. Investigation into the relationship between pheromone

attractiveness and success in mating trials.
Behavioural measures of courtship:
The nature of the design of this experiment allows us to investigate whether
differences in pushing behaviour are associated with success in being the first of a pair of
91
m ales to mate with a fem ale. To this end, behavioural data for replicates where a mating
occurred were divided up into those performed by the mated and those performed by the
unmated male. A ll encounters observed were categorized according to which beetle initiated
the encounter (w ho approached whom) and what the result o f that encounter was (male
push; fem ale push or no obvious push). Each category o f encounters was then averaged
across trials and plotted as a bar chart (see figs. 6.5.5.a and b). Averaging proportions
instead o f absolute frequencies enables each trial to contribute equally to the chart.
0.4 □ no push
□ female push
amale push
0.3
o 0.2
£ 0.1
ts
4ffa3t
' ::
o
2
CL,
Fig. 6.5.5a: First set of trials (N=27). See legend below.
0.6
£o2 0 .5
c3
oo
c<D 0 .4
2o
0.3
C
m
o
c
o 0. 2
c
a2
CL 1
0
Female Mated male Female Unmated
approached approached approached male
mated male female unmated approached
male female
Fig. 6.5.5b: Second set of trials (N=34).

Fig. 6.5.5a and b: Bar charts to show the proportion of encounters leading to three different behaviours
(male push; female push; no obvious push) for all encounters between males and females in a mating arena.
Two males were presented to a single female and the males have been subsequently categorised according to
whether they were the first of the pair to mate with the female. Encounters have been divided up according
to which beetle initiated the encounter (who approached whom).
92
Sex differences in behaviour are clearly shown in figs. 6.5.5 a and b. Females
are more pushy than males as found in section 6.5.4. No consistent difference between
male and female approach rate has been found: females initiated more encounters in the first
trials and males initiated more encounters in the second trials.
Raw counts of each behaviour scored were compared between the mated and
the unmated males using a Wilcoxon signed-ranks test (the data do not approximate to a
normal distribution). Only the number of encounters where the male approached the female
and no push ensued was significantly different between mated and unmated males of a pair:
1st trial, N=27, z=-3.39, p=0.001; 2nd trial, N=34, z=-4.59, p=«0.001. However, the
difference in the proportion of encounters that resulted in no push where the male
approached the female for mated and unmated males was just verging on significance
(using a Wilcoxon signed-ranks for replicates where both males approached at least once:
N=37, z=-1.79, p=0.07). This low significance may be due to error incurred from
calculating proportions from low frequencies of observations. Still, it is more accurate to
say that the data support the hypothesis that males who approach females most frequently
are more likely to secure the first mating, and no definite conclusions can be made about the
likelihood that these encounters will result in a female push.
The number of times females approached males was not found to be

significantly associated with winning a mating contest (Wilcoxon signed-ranks, N=61,
z=0.384, p=0.70). However, this variable might indicate whether females continue to
orientate to aggregation pheromone at the distances experienced in the mating arena. Males
were therefore re-categorised into those who had scored highest and lowest in the volatile
bioassay. The total number of encounters where the female approached the male was then
compared between these two groups using a Wilcoxon signed-ranks test. This difference
was not statistically significant (N=66, z=-1.69, p=0.09).
The relationship between mate choice, volatile preference and

relative male weight:
Frequency data describing the associations between mate choice, volatile
preference and relative male weight are presented as a series of tables below (see tables
6.5.5c-e). In the 81 replicates tested, no statistically significant associations were recorded
between these three variables. Larger males were chosen first more frequently in mating
trials and visited more often in volatile bioassay trials. This is consistent with the tendency
recorded in chapter 4 (section 4.4.4) where larger males gained higher bioassay scores.
However, even combining these results does not produce a significant association between
male weight and bioassay performance.
93
Tables 6.5.5 c, d and e: The association between males preferred in pheromone bioassays, those that mate
first in mating trials, and male size.
C. Volatile preference vs. success in mating trial:
r -- - - - - - - - - - - *~ \
I Preferred volatile = | Preferred volatile * | No preference
j
! |
\
First to mate i First to mate 1I
i
1 1st Trials |I 15 1 8 1 13
1!. . . . . . . . . . . . . . . ■. . . . . . . 1I ■■
j 2nd Trials I 10 I\ 17 {
\
18
1
1 Total 1I 25 1 25
J.......................................................1
| 31
i
D. Volatile preference vs. relative weight o f males:

.. i ........
| Preferred volatile = ( Preferred volatile ^ No preference
| | Heavier male Heavier male [ j
J ......................................................j
j 1st Trials
1 ................................... 1 21 I| 11 ^I 4 1
1 2nd Trials 1 18 1 16 1i 11
j
1................... !
I Total | 39 11 27 1i 15 I
»LU t.................. ...twnm
wwwuwm
mw^M iM
ii
E. Success in mating trial vs. relative weight o f male:

i
| First to mate = | First to mate * | No preference I
£ j Heavier male i Heavier male ( I
\ \
* I
j£ 1st Trials 1 16 I1 11 1\ 9 |
| 2nd Trials I 22
I 1 12 i
.1..................... \.... 12 |
I|
i
1\ I 1
u>
Total 23 [ 21
0°
Absolute weight o f males, volatile preference and behavioural measures o f courtship:

Pairs of males of fixed weight difference have been randomly sampled from the
range of weights found in a culture of LGB. This means that it is possible to ask the
question, does the average weight of a pair of males influence the total response and/or
pattern of preference for the larger or smaller male?
Absolute weight o f male vs. volatile response/preference:

Total bioassay score was not found to vary significantly with average weight of
the males presented to the female. Bioassay score given to each male was also found to
vary randomly with individual male weight.
The bioassay response given to the larger male minus that given to the smaller
male was plotted against average weight of the males. Trials 1 and trials 2 were plotted
separately in an attempt to visualise any consistencies in trends in the data. A moving
94
average of 10 replicates was used to smooth the data. Neither a consistent direction of size
preference nor an optimum size preference model are supported by the data.
Absolute weight o f male vs. measures o f courtship behaviour:

No significant differences were found between the behaviour of the larger and
smaller male of a pair. However, there is some indication that the heavier male of a pair is
more likely to win a mating contest. We tested behaviour variables against the average
weight of the pair of males to see if they were correlated.
The proportion of encounters in a trial classed as ‘male approach, no push’ was

not significantly correlated to the average weight of the two males used in the trial:
Spearman’s rank N=46, rs=0.173, p>0.1 (see fig. 6.5.5f). This variable was tested since
this was most correlated to mating success when males within a pair were compared. There
is some suggestion of two populations of points in the data. Closer examination of the data
revealed that this apparent separation is not due to the data being taken from two sets of
trials. No variable measured has been found to account for this apparent pattern.
0 .9 -
c
3
o0 —
0.8
5§ 0 .7
2 O
Be
j£ 0. 6 -
1ee oa 0 .5
s s.
3o ea)o* 0 .4 -
ci_ 15
° B 0 .3 -
G ii
0 II
'£ 0. 2 -
••
C1
u
••
0.1
• •%
2 .5 3 3 .5 4 4 .5
Average male weight (mg)
Fig. 6.5.5f: Graph to show the proportion of total male/female encounters arising from a male approaching
the female that resulted in no push, against the average weight of the pair of males presented (Spearman’s
rank N=46, rs=0.173, p>0.1).
Body size could conceivably influence the likelihood that females may push a
male, however proportion of encounters resulting in a female push was not correlated to
average body weight of the pair of males. Another possibility is that the number of
encounters prior to mating may be correlated to body size, this was also found not to be the
case.
95
Male-male interactions:
21% of all interactions recorded during mating trials were between the pair of
males (Trial 1 (179/911+179); Trial 2 (326/992+326). In the first trials 30% of male-male
interactions resulted in an obvious push compared with 34% in the second set of trials.
Males are less aggressive towards females, the percentage of male-female interactions that
resulted in the male pushing the female was 3.5% (first trials) and 5% (second trials).
The influence of male size on male-male interactions was tested. Larger males
were pushier in both sets of trials but this difference was only statistically significant in trial
1 (see table 6.5.5g).
Table 6.5.5g: The average number of pushes for smaller and larger males of a pair during male-male
encounters for two sets of similar trials.
.................................................................................................................................................................................................... |*....................iTniir‘"'rri‘nnrT^iiTmTiTTrnnriinmr^TrTiwrririiniTfi¥niiinmrrnT[rfiinAmaiiiiAMaw!<,W-Wt!ti!**(i-!;^
| Average number of pushes per || I

I trial I| I
... -------------------- ------------ ---- —^----
j Larger male 1I Smaller male 1i Test stat. 1 N I
! I
| 1st Trials ) 1.11 ) 0.39 1 z=-2.20 | 36
|i 1$ 1
| 1 j 1
1 p=0.03 |
.1.... ........... |... ......... J. i
| 2nd Trials 1 1.80 !
1
0.67
\
I z=-1.59 | 45 1
1 ||
! I | p=0.11 |
1 I 1 1
To investigate the possibility that the outcome of male-male interactions is

correlated to the outcome of male/female interactions, the data were divided into pushes
performed by the male who subsequently secured the first mating and pushes performed by
the other, losing male (trials where no mating occurred were eliminated from this analysis).
Winners of mating trials were significantly more pushy in male/male interactions in the
second set of trials, no difference was found in the first set of trials (see table 6.5.3h).
Table 6.5.3h: The average number of pushes given by winners and losers of mating trials to each other
during mating trials. Data from two sets of similar trials are presented.
|-----------------------
Mated male I Unmated male 1 Test stat. ] N
ji . 1. . . . . . . . . . . . . . . . . . . . . . 1 1
| 1st Trials I 0.52 1\ 0.56 11 z= 0.07 1 27 1
% I p=0.94 NS |
| j |
| 2nd Trials | 1.68 | 0.45 | z=-2.48 j 38 j
I 1
| p=0.013 |
96
6.5.6 The influence of manipulating pheromone signal on courtship

behaviour.
Pheromone bioassay trials
Males in the fresh grain group were at least an order of magnitude more
attractive to females in bioassay trials than males kept in the Female-Factor treatment
(average response of 3.13 visits per trial compared to 0.2 visits per trial). A Wilcoxon
signed-ranks test between each pair of males revealed that this difference was significant,
z=2.809, N=30, p<0.01.
Mating trials
No significant effect of treatment with Female Factor was recorded on overall
courtship success (see table 6.5.6a) or on specific measures of pushing behaviour during
courtship (see behaviour measures section below). In fact, the data are more notable for
their consistencies between observations made before and after treatment than for any
demonstration of effect of the treatment. The same male mated first in both trials conducted
on each replicate significantly more times than not (N=23, chi-squared = 9.78, p<0.01)
(see table 6.5.6b). This shows that mating preferences are consistent between females since
different females were used before and after treatment. Even though no significant effect of
treatment was demonstrated, in all four cases where the first male to mate changed after
treatment, the switch was from the untreated to the treated with Female Factor male.
Table 6.5.6a: The association between treatment with Female Factor and success in being the male to mate
first in mating trials.
I Female Factor male mated | Untreated male mated first j No mating j

f first j
1......................................................
|
I
14 j
\
12 |
I
4 j
Table 6.5.6b: Table to show the numbers of replicates where the same male was the first to mate in trials
before and after treatment.
1................................. ...........
Consistent result | Inconsistent result | Cases where no result one j
} or both trials j
I 19 | 4 1 7 1
& .................. .... 1
Behavioural measures (encounter data):

Behavioural measures are likely to be more sensitive to changes induced by the
treatment than ultimate mate choice. Therefore, the behaviour highlighted as that most
closely associated with mating success (number of times each male approached the female
resulting in no push) was examined for treatment effect. Since male pairs were kept
constant before and after one of the males was treated with Female Factor, influence of the
treatment could be accurately measured by comparing the pair of males before and after
97
treatment. The measure of the relative amount each of a pair of males pursued the female
used is shown below:
(No. approaches bv Treated male = no push) - (No. app.bv Untreated male = no push)
Total number of times the female was approached by a male
A Wilcoxon signed-ranks test was used to test if there was a consistent

difference in the relative behaviour of pairs of males before and after treatment. No
difference was found, N=30, z=1.91, p=0.056, but the trend indicated is forFemale-
Factor treated males to pursue females more successfully. This measure of behaviour was
surprisingly consistent before and after treatment (Spearmans rank: N=30, rs = 0.61,
p<0.005) (see fig. 6.5.4 c).
Earlier results had indicated a trend for females to approach the male they had
preferred in a bioassay trial more than the other male in a mating trial. However, when the
pheromone signal was manipulated in this experiment, females did not then orientate less
often to the males who had shut down their pheromone signal (those treated with Female
Factor) (Wilcoxon signed-ranks test for replicates where females approached at least once
in each trial: N=14, z= 1.269, p=0.20).
- 1
Before treatment
Fig. 6.5.4c: Graph to show the relative amount each male of a pair approached the female and no push
ensued, before and after one of the males was treated with Female Factor (Spearman’s rank: N=30, rs = 0.61,
p<0.005).
98
Male-male interactions
No effect of treatment with Female Factor on male-male pushing behaviour
was found. The proportion of male-male encounters resulting in a push remained
approximately the same (mean before treatment = 0.33, mean after treatment of one male =
0.31, Wilcoxon signed-ranks test of replicates where males had at least one encounter in
trials before and after treatment: N=26, z=-0.143, p=0.89). The relative pushiness of each
male of a pair was also unaffected by treatment of one of the males with Female Factor. The
relative pushiness was calculated as:
Number of pushes bv male to be treated-No. pushes bv untreated male

Total number of pushes between the two males
Only seven replicates were suitable for this comparison since there were many
instances where no pushes were recorded between the pair of males and where there was
no change in the relative pushiness with treatment. Of the seven replicates where the relative
pushiness changed after treatment, treated males became more pushy in four replicates and
less pushy in three replicates.
6.6 DISCUSSION
Variation in courtship and copulation behaviour between species is
bewilderingly vast (see Eberhard, 1996 for a review). Therefore an initial description of
such behaviours can very quickly rule out the possibility of some of the more extreme
behaviours found in other species, and suggest avenues for further study. Outside the plant
host, LGB courtship behaviour is characterised by a particularly variable pre-copulatory
period of pushing behaviour, a period of stereotyped rubbing of the female using male
antennae and legs, a short copulation, and little evidence of any post copulatory interaction
between male and female. This is in contrast to behaviour within tunnel systems, where
male and female pairs were found to remain together, in some cases probably for the entire
observation period (5-6 days), and to mate repeatedly during this time. In these preliminary
observations only one female was observed to co-habit with each male, unlike many
polygamous Scolytid species (Kirkendall, 1983; Schmitz, 1972).
Fadamiro’s observation that males are, ‘the mate finding sex’, is not
convincingly supported by the data presented in this chapter. In open arenas and within the
plant host, both male’s and female’s movements bring pairs together for copulation.
General reluctance to mate by females, observed as pushing behaviour, results in female
choice between potential mates. Stereotyped rubbing between mates as described in phases
two and three of courtship (6.3) (usually performed by the male on the female) is
widespread in insect courtship (Eberhard, 1996). Copulation duration in LGB is very short
whether insects mate in or outside a plant host. This contrasts with that found in Sitophilus
oryzae (Holloway and Smith, 1987), and in the milkweed leaf beetle (Dickinson, 1988);
99
Leaf beetles remained coupled for an average of 0.75 days in Dickinson’s study.
Presumably the female constrains sperm uptake and or displacement rates, or promotes an
extended period of sperm delivery by some other form of cryptic female choice in this leaf
beetle. This is predicted since risk of interruption to mating should otherwise favour males
who deliver their ejaculate as fast as possible. There is little evidence that Prostephanus
truncatus males continue to court females after intromission begins. In fact both male and
female are remarkably still during this phase and the pair part company directly after this.
Continued courtship after intromission has begun can indicate that males can benefit from
continued stimulation of the female presumably by influencing variables of cryptic female
choice (Eberhard, 1996). Therefore cryptic female choice is not suggested by the
observable copulation behaviour of this species.
Two measures of readiness to mate, time before mating, and number of

encounters before mating, were found to be variable and characteristic of which individuals
were involved in the mating trial. Number of encounters before mating is perhaps a better
measure of readiness to mate than time before mating, since time before mating includes
periods when beetles are relatively inactive which can unduly influence the data. Female
identity was a much greater influence than male identity on readiness to mate. The effect of
variation in female behaviour has not been investigated further in this study as differences
between males are the focus here, but female variation in stubbomess to mate (level of
choice) would be an interesting line of investigation.
Different females were found to mate preferentially with the same males when
given a choice of two different males. Neither male variation in aggregation pheromone
signal nor fresh weight were found to be clear determinants of the non-random pattern of
mating observed in LGB. Manipulation of the pheromone signal confirmed that this signal
does not greatly influence mate choice during contact courtship. Pheromone manipulations
also did not influence pushing behaviour between males. However the sensitive nature of
the experimental design did pick up a light trend for males treated with Female Factor to
increase the behavioural measure most correlated with mating success (male approach, no
push). Therefore, although the aggregation pheromone signal is not used by females to
distinguish between mates at close quarters, the female produced pheromone, Female
Factor, which is already known to alter male aggregation pheromone emission, may also
induce behavioural changes in males.
Male size is correlated with pushing behaviour between males, with the larger
male of a pair being more likely to push the smaller male than vice-versa. Such behaviour is
similar to that reported for Monochamus scutellatus (Cerambycidae) (Hughes, 1981),
where larger males win more often, but males do not apparently actually damage each
other. Direct male-male conflict is reviewed for Scolytidae in Kirkendall (1983). Pushing
behaviour such as that described here for LGB is also a feature of Scolytid behaviour.
100
Pushing between male Scolytids has been observed on the bark surface. Generally it is
noted that once a male is within a gallery he cannot be forced out (Oester and Rudinsky,
1975, quoted in Kirkendall, 1983). Size differences were not found to influence the largely
age dependent aggressive interactions of the cockroach Nauphoeta cinerea (Moore et al.,
1997).
These results indicate that all mate acquisition benefits from pheromone
signalling occur from manipulating female distribution. It has already been shown that
pheromone perception of distributing beetles in the field has a high resolution, and beetles
can distinguish between two male signals placed 0.6m apart. Therefore aggregation
signalling is likely to be a major determinant of mating success when males occupy distinct
tunnel systems. However the pheromone signal does not apparently influence mate choice
between two or more males occupying the same system. This means that males employing
a satellite strategy of following other male’s signals may successfully mate with any
females they intercept even if they are not signalling (which is likely if they are not
feeding). However, once females enter a signalling male’s tunnel system she becomes
relatively easy to guard, which could limit the risk of lost paternity to satellite males.
If LGB has only recently expanded its host range to include stored products
and is indeed a relatively poorly adapted storage pest (Hodges pers. comm.) then LGB
tunnel systems may be poorly adapted to the limited space of a maize grain. The long term
association between pairs of LGB observed in artificial hosts may break down in maize
since first, there is a greater risk that tunnels will break through to the surface of the grain,
(although there is evidence that LGB can detect the thickness of the medium it is tunnelling
and may stop in time) and second, other beetles can break into the tunnel system from all
directions, which is not possible for a much larger host.
Mate guarding in LGB appears to be flexible and dependent on the location of

mating. Couples apparently only remain together for any appreciable length of time when
mating occurs in the context of tunnel systems in a plant host. It may only be feasible for
males to attempt to monopolise a female when she is inside a tunnel that limits her
movements, and is easy to defend. LGB behaviour within tunnel systems is startlingly
similar to that described for some Scolytid species. Kirkendall notes of Scolytids that,
‘..males that stay during gallery construction remain blocking the entrance hole,
occasionally leaving their post to copulate, feed, or remove accumulations of frass..’
(Kirkendall, 1983).
Male LGB guard females by the fourth method described by Alcock (1994);
‘monitor a mate without physically grasping her following completed copulation,’ and
possibly also by the second method; ‘donate mating plugs after insemination,’ since the
spermatophore may act at least temporarily as a mating plug. Post copulatory guarding can
101
also be seen as pre-copulatory guarding since male LGB often mate repeatedly with the
same female. Male LGB presumably incur costs of staying with the female from lost time
searching either physically or through pheromone signalling (since the signal is shut down
in the presence of a female). The original starting hypothesis for this thesis proposed that
males shut down the pheromone signal to limit male-male competition. It is, however,
possible that on an evolutionary time scale, females could potentially force males to
continue signalling by preferentially mating with signalling males. This idea is similar to
that discussed in Andersson (1994) of, ‘female incitation of male competition’. In the
elephant seal, Mirounga angustirostris, females protest loudly against being mounted
especially if the male is young, leading to attraction of competing males (Trivers, 1972
quoted in Andersson, 1994). Therefore the question, ‘should males shut down the
pheromone signal in the presence of females?’, can be viewed from both a male and female
perspective and as such, the answer will represent the outcome of potential conflicts and
concordance of interests of males and females. I have summarised the main pros and cons
of continued signalling below:
Possible selection pressures acting on MALES to continue

signalling in the presence of a female
• Increase the number of potential mates.
Possible selection pressures acting on FEMALES to force male to

continue signalling whilst they are together
• Promote male-male competition from the arrival of males responding to the
pheromone signal.
Possible selection pressures acting on MALES to cease signalling

in the presence of a female
• Decrease the potential for male-male competition from other males
following the pheromone signal.
• Decrease the level of predation of possibly adults, but more likely young
offspring, from predators that follow the aggregation pheromone signal to
locate their prey (pheromone acting as a kairomone).
• Energetic/nutritional gains if signalling is costly in these respects.
Possible selection pressures acting on FEMALES to either favour

males who shut down their signal or at least be unprejudiced
towards non-signalling mates.
102
• Decrease competition for oviposition site from females following the

pheromone signal.
• Decrease the level of predation from predators following the aggregation

pheromone signal.
Why, then, do females not continue to use the pheromone signal to choose
between mates? Perhaps the answer is that natural selection has favoured males that shut
down their signal in the presence of ovipositing females in order to limit predation. If this
was an important selection pressure then females who continued to remain with such males
would also be favoured and mate selection during courtship on the basis of pheromone
signal would not be promoted. Blocking the entrance to a gallery system may not only limit
conspecific access, it may preclude access to parasites and other predators. Some evidence
that male presence increases numbers of live offspring produced through decreased
predation in the Scolytidae is cited in Kirkendall (1983).
The alternative explanation is that males cease signalling to limit attraction of

other males; this assumes that females, who could benefit from increasing the level of male-
male competition, cannot reverse this trend by continuing to select for signalling males.
This scenario is further complicated by the fact that as long as males signal, extra females
are also attracted. Depending on the size of the food resource, competition for oviposition
sites may also contribute to the benefits incurred from cessation of signalling. Alternatively,
it may just be that pheromone signals are difficult to assess at very short range. Volatile
chemical signals may merge with each other and may not be a reliable source of
information. However, Moore’s studies of cockroaches would tend to indicate that this
need not be the case (Moore 1998).
Surprisingly, previous mating experience (in terms of number of matings

already performed up to six possible matings) was not found to be correlated with readiness
to mate (experiment 6.4.4). Thus no reduction in willingness to mate of either males or
females was demonstrated for the average recovery time of 12 hours used in this study.
Although transfer of sperm was not directly confirmed in these trials, later observations
have shown that phase 4 of courtship is a reliable indicator of sperm transfer. Observations
of beetles tunnelling in the plant host confirmed that LGB adults mate repeatedly
(commonly once or twice each 12 hour light period). These observations are more reliable
indicators of possible mating rates in the wild since beetles have the opportunity for other
behaviours such as feeding and oviposition.
LGB females were found to mate repeatedly with the same partner. Petrie
(1992) discussed multiple mating of the same partner in birds and proposed three possible
103
adaptive explanations: 1) to ensure fertilization (not much evidence for this in birds); 2) to
increase paternity assurance for the male such that the female might gain from consequently
increased parental care; 3) to decrease the probability that the male will mate with other
females and therefore the female will keep good territory associated with the male.
Dewsbury (1982) cites costly ejaculates as a possible reason for multiple mating with the
same mate, “...because of the stimulus requirements for pregnancy initiation, sperm
competition, female choice and control, and the costs and risks of searching, the males of
many species may be selected to copulate repeatedly with a single female”. Dewsbury
proposes that the balance in this optimality theory depends on the operational sex ratio.
Multiple mating with the same partner may represent a form of male parental investment.
Fox (1993) suggested that benefits incurred by female Callosobruchus maculatus
(Coleoptera) are largely through nutritional benefits from the large ejaculate delivered by the
male (but see some limitations discussed in Fox et al., (1995)). Multiple mating thus
suggests that sperm competition, cryptic female choice and/or male paternal investment are
possibly important determinants of variation in male reproductive success in Prostephanus
truncatus. The next chapter investigates some of these phenomena.
104
Chapter 7: Sperm competition
CHAPTER 7: SPERM COMPETITION
7.1 INTRODUCTION
The influence of pheromone-directed dispersal and courtship behaviour on
mate selection in Prostephanus truncatus has already been considered in chapters 4-6. This
chapter touches on the next step that determines relative male reproductive success:
processes that occur between the onset of copulation and the fertilization of eggs.
Prostephanus truncatus is likely to be a good subject for the study of sexual selection
within the female tract since it has already been shown that both males and females will
mate several times. Also, males and females are likely to encounter many possible mates
during their relatively long adult lifetime, especially given the aggregation behaviour of this
insect. It has long been recognised that both inter- and intra-sexual selection are possible
within the female tract in the form of sperm competition and cryptic female choice (Parker
1970). Sperm competition is defined by Parker (1970) as, “the competition within a single
female between the sperm from two or more males for the fertilization of the ova”. Cryptic
female choice is a term used to describe, “female processes that affect male reproductive
success and occur after the male has succeeded in coupling his genitalia with those of the
female”, (quote from Eberhard, 1996). Thus the race for fertilization is a battle between
male ejaculates whose outcome can be influenced by preferential treatment of some
ejaculates relative to others by the female.
Reproductive morphology/chemistry can tell us an enormous amount about the

likely determinants of fertilization success between males in polyandrous species like LGB.
Reproductive structures are notoriously diverse, complex, and often surprisingly extreme in
form. Extreme forms are likely to have resulted from sexual selection (Andersson, 1994)
and can therefore indicate fruitful areas for further study. For these reasons the reproductive
organs of male and female LGB were dissected, described and discussed in this chapter.
Scholz (1997) has since given a basic description of the reproductive organs of
Prostephanus truncatus, I will therefore concentrate on reporting details of those parts of
the reproductive tract which could influence/have been influenced by sexual selection.
The proportion of offspring sired by the second male to mate in a double mated
female (P2value) is a useful parameter often measured in studies of sperm competition. It is
unusual for the first male to mate to secure 100% of paternity if the female remates (Ridley,
1989). Last-male sperm precedence is common in insects. The extent to which males can
secure paternity of offspring from multiply mated females is interesting to the discussion, of
the potential costs incurred by male signallers who attract other male competitors. Paternity
can be traced using genetic markers, or the so called ‘sterile male technique’ (Eady, 1991).
105
No genetic markers currently exist for LGB so the sterile male technique was attempted in
this study. In this method one of a pair of males allowed to mate with a female is sterilised.
Eggs laid are then scored as viable or non-viable. Careful control of mating order allows
possible influences of differing fertilization success between sterile and non-sterile males to
be eliminated from the calculation of P2 (see Boorman and Parker, 1976, quoted in Eady,
1991).
One of the most striking features of patterns of sperm/ejaculate investment is

that it is extremely variable, both between individuals (Lewis and Austad, 1994; Simmons
and Parker, 1992); for the same individuals of different ages (Fox et a/., 1995);
physiological condition (Simmons and Parker, 1992); and, facultatively, in response to
social cues (Gage, 1995; He and Miyata, 1997). Previous work has found that males
sometimes alter the size of their ejaculates in concordance with predictions derived from
sexual-selection theory (Gage, 1991 in a fruit fly). For instance, male Mediterranean fruit
flies were found to increase their sperm number per ejaculate by almost three fold in
response to increased levels of perceived male competition (presence of another male)
(Gage, 1991). Siva-Jothy and Tsubaki (1989, cited in Eberhard 1996), found that male
damselflies of Mnais pruinosa delivered less sperm per ejaculate when they were territorial
compared to when they were non-territorial. Female damselflies tend to oviposit directly
after mating with territorial males (thought to be before sperm mixing becomes important)
and territorial males often have other possible females to mate soon afterwards, which
could limit the number of sperm they have available. Therefore there are two possible
differences in the selection pressures on the size of ejaculate in territorial vs. non territorial
male damselflies that could explain this observed difference in ejaculate size.
Such short term facultative adjustment of ejaculates might allow us to determine

why ejaculates are the size they are, and what cues males are responding to when altering
their ejaculate size. This chapter investigates whether LGB males alter the sperm number in
their ejaculates in response to male crowding such as that found by Gage (1995), and also
in response to Female Factor (see chapter 3 for a description of Female Factor). It has
already been shown that LGB males use Female Factor as a cue to decrease their level of
pheromone signalling. Thus, LGB males already use Female Factor as a source of
information about their current sociosexual environment and adjust their pheromone output
accordingly, so it is perfectly feasible that this might be a cue that triggers facultative
adjustment of sperm investment.
It was initially envisaged that males might invest more in ejaculate production
when exposed to Female Factor as it has been proposed that amount of Female Factor could
be used as a measure of local population density (Fadamiro, 1995). Total pheromone shut
down takes approximately six days (chapter 3) and initially, males were not observed to
stay with females after mating (in open arenas, chapter 6). Exposure to Female Factor was
106
therefore likely to be determined by the local population density and not so much by the
chance association of males with a single female. Observations made using time-lapse
recordings of LGB behaviour in tunnels within the plant host, however, show that males
do in fact co-habit in tunnels with the same female for extended lengths of time and so a
single female could trigger pheromone shut down in a male. Apart from increasing our
understanding of the mating system in LGB, being able to trigger adjustment in sperm
investment using a chemical cue could be a useful tool for more general studies of sperm
competition and cryptic female choice.
Ejaculate and sperm investment are broad terms that encompass a number of
variables. This study measured sperm number per ejaculate and approximate ejaculate size
and weight since these measures are relatively easy to obtain and have already been shown
to be variable. Ejaculates usually contain many other substances as well as sperm (see
chapter 6 on male sexual products in Eberhard, 1996), which may alter in proportion
between ejaculates. Another measure of ejaculate investment is speed to mating, since males
may continue to deliver full ejaculates, but may mate at a lower rate when overall
investment in ejaculates is lower. Females are often assumed to be limited in the number of
eggs they can produce. This is suggested for LGB since eggs are resorbed by females in
times of food shortage (Scholz, 1997). Contrary to early ideas, males have also often been
demonstrated to be limited in their gamete production. Each sperm cell may indeed usually
be relatively cost free, but ejaculates as a whole, which often contain thousands of sperm
and other materials, may be costly to produce (Fox et al., 1995 in a beetle; and see
references cited in Dewsbury, 1982). Differences in ejaculate investment may only become
apparent if males are given the opportunity to mate twice or more in fairly quick succession.
Comparing first and second ejaculate sizes and readiness to mate, can also show whether
males are at all limited in ejaculate constituents and how they partition investment (full
ejaculates less often vs. smaller ejaculates at the same rate).
To summarize, the main questions are: do the first males to mate females risk
losing paternity if females remate?; and do males adjust their investment in sperm in
response to such a threat? It has already been shown in chapter 6 that females do not use the
aggregation pheromone signal in inter-sexual selection for mates during courtship, so males
who cease signalling should not decrease their reproductive success at this point. Whether
female Prostephanus truncatus show cryptic female choice that is influenced by the
pheromone signal is unknown and still a possibility.
7.2 AIMS
• Make a preliminary assessment of the potential for inter and intra-sexual selection at the
level of the gamete in Prostephanus truncatus from dissection and description of
reproductive organs.
107
• Describe the pattern of sperm precedence in doubly mated females.
• Determine if males manipulate their investment in ejaculates in response to the presence

of other males or the female produced pheromone, Female Factor.
7.3 METHODS
7.3.1 Dissection of reproductive organs
Reproductive organs of male and female Prostephanus truncatus were
dissected out under a stereo microscope (Nikon model SMZ-1) using fine forceps.
Specimens were mostly dissected floating free in a drop of insect buffer as specimens were
generally too small to pin into wax.
7.3.2 Sterilization of males for calculation of P2

Source o f insects
All insects used were removed from 4-5 week old cultures as cased pupae.
Virgin females were obtained by isolating pupae before hatching.
Apparatus
A 5 mv linear accelerator (X-ray machine; Medical Physics department,
Leicester Royal Infirmary) was used to administer controlled doses of radiation. Pupae
were housed in a 5 x 5 divided petri dish, with one pupa per division. Perspex blocks
provided the appropriate build up and scatter of the incident radiation.
Experimental design
In the course of two separate sets of trials, cased pupae were exposed to a
range of radiation doses: 120GY, 60GY, 32GY, 30GY, 15GY, 8GY, 2GY and OGY
controls. 25 pupae were exposed to each dose (approximately half males and half females).
All surviving males hatched from these pupae were allowed to mate with virgin females on
split grain. Grain was subsequently dissected and all offspring produced were recorded.
7.3.3 Determination of sperm number per ejaculate for two

consecutive matings of males kept in different sociosexual
environments.
Preliminary observations showed that male LGB are capable of producing two
spermatophores within five hours. A time of five hours between double matings was
chosen since this allowed one batch of insects to be tested in a nine hour working day.
All dissections were initially carried out in an insect buffer solution as

described in Gage and Cook (1994). Buffered solutions reduce damage to sperm structure
during dissection and dispersion. However, using buffered solutions incurs the
disadvantage that sperm can sometimes still be obscured by salt crystals even after
washing. Distilled water was therefore used to disperse ejaculates in some experiments.
108
LGB sperm can become damaged during this process (axoneme unravels), but a very clear
sample is obtained. Possible overestimation of sperm number by counting the two strands
of the tail as separate sperm was prevented by limiting sperm counts to those with a head.
Source o f insects
All insects used in trials were recently hatched adults (2-5 weeks old).
Tunnelled kernels with no adults were taken from 30 day old culture jars. These were then
placed in a fresh culture jar and left to incubate for two weeks. Each wave of insects used
was sieved from such a culture, sexed and placed into treatment pots. Only fully
sclerotised, active insects were used (likely to be at least five days old). Therefore insects
used were unlikely to be virgin, but were mostly of a similar age.
Apparatus
The mating arena consisted of a petri dish base lined with paper towel with a
clear plastic pot placed upside down on the towel. Insects were introduced under the plastic
pot. This provided a circular walking arena of diameter 3cm. This is a much smaller arena
than that used for the mating trials conducted in chapter 6 (diameter 9.5cm). It was hoped
that this would increase the encounter rate between the male and female and subsequently
decrease the time before mating, yet still allowing courtship behaviour to proceed.
Sperm-count protocol
Ejaculates were dissected out of recently mated females. Intact spermatophores
were often expelled from the female during dissection. If not then they were dissected
straight out of the bursa copulatrix. The bursa was always checked for sperm that might
have been expelled from the spermatophore. When the spermatophore burst or no
spermatophore was found, all clumps of sperm were retrieved.
All dissections and dispersions were carried out in distilled water. Debris was
cleared away from the sperm sample leaving it in a clear drop of water. The sperm were
then dispersed in this drop by teasing apart clumps of sperm using fine forceps and then
gently mixing the sperm into the drop. When no large clumps of sperm remained, the
ejaculate was then made up to 15ml with distilled water by using this water to wash the
sample into a 25ml beaker. The sample was then mixed for four minutes using a blunt metal
rod, in the beaker. Four 20|il samples of this mixture were then placed as drops on a glass
slide using a 20|il micropipette (Gilson). The mixture was stirred directly before each
sample was taken and each sample was taken from a point half way up the sample and half
way from the center to the outside of the beaker to limit any influence of sperm settling
through the sample.
Drops were left to dry under a dust cover. Sperm were examined using a
compound microscope set to dark phase contrast under a magnification of x8xl0 (Olympus
109
BH-2). The numbers of sperm per 20|ll1 drop were counted by systematically scanning
down each drop. An estimation of the total number of sperm per ejaculate was then
calculated by multiplying up the average value per sample x 750 (15/0.02ml). By taking
more than one sample from the diluted ejaculate an estimation of the sampling error was
made. All implements were rinsed after use and new pipette tips were used for each
ejaculate sample.
Experimental design
All sexed females were placed in individual glass pots containing fresh ground
maize meal. All males were randomly allocated to one of three treatments:
• Single, fresh: Single male in pot containing fresh ground maize meal.
• Single, conditioned: Single male in pot containing one part fresh

ground maize meal and two parts flour taken from a 6 week old culture.
• Crowded, fresh: X10 beetles per pot containing 5cm3 of fresh ground
maize meal.
Insects were kept in treatment pots for 7-10 days before being used in trials.
Four cohorts of insects were used in successive trials. For each replicate a male
was given the chance to mate with a female. As soon as they separated after mating the
female was removed, the ejaculate was dissected out and the number of sperm estimated. If
no mating occurred within 25 minutes the trial was abandoned. Males that mated were then
placed back on food and left for five hours. After this time they were given the opportunity
to mate with another female. Again, if they mated within 25 minutes, the female was
dissected and the sperm number per ejaculate was estimated. The sizes of all intact
spermatophores were measured under a compound microscope. Generally six males were
tested each day, two from each treatment. The order of trials was changed each day to
eliminate time of day effects from the treatments. A total of 62 such trials across the three
treatments were performed. For each trial the following variables were measured:
• Time taken to mate (or if mating occured)
• Spermatophore length and width at the widest point (if spermatophore still
intact)
• Sperm number delivered in the ejaculate
• Male weight one day after the trial (In the last wave of trials male and
female weight before and after each mating were measured instead)
110
Chapter 7: Sperm competition.
Additional controls:
In successive mating trials, all first matings were performed in the morning and
all second matings were performed in the afternoon. 21 additional trials were therefore set
up where first matings were performed in the afternoon to check for time-of-day effects.
Equal numbers of males used in these trials were taken from each treatment category.
To estimate the number of residual sperm, five females selected randomly from
the bank of those being used in trials were dissected for sperm without being allowed to
mate with a male. The spermatheca of these females were purposely opened up to obtain an
upper estimate of the number of sperm that could contaminate ejaculate samples.
7.4 RESULTS
7.4.1 Dissection of reproductive organs
Female reproductive organs
Mature females of LGB possess one pair of ovaries. Each ovary has five
ovarioles (only three are shown in fig 7.4.1a, for clarity). Ovarioles contain a string of up
to about four eggs at various stages of development with the largest and most mature
attached to the ovary base (see fig 7.4.1a). The oviducts are lined with muscle tissue (fig.
7.4. lb) and they open into a vagina enlarged to form a bursa copulatrix. Also leading off
the bursa copulatrix is a duct leading to a spermatheca (sperm storage organ). The bursa
copulatrix is continuous with the outer body wall and terminates as a yellow sac, possibly
a store of fatty tissue and/or of secretory function at the point where eggs are squeezed out
of the body.
The spermatheca consists of a spherical sperm reservoir (receptaculum

seminis), a leaf shaped accessory gland and a blind-ending duct (see fig. 7.4.1c). The
sperm reservoir is sclerotised and spherical in shape. Some contractions were observed
around the blind-ending ducts after some dissections, however the exact structure of this
muscle tissue was not discernible. Dissection of reproductively active females revealed the
sperm reservoir to be full of active sperm, but no such sperm were found in the blind
ending duct. In some dissections there appears to be a broken connection between the end
of the blind ending duct and the sperm reservoir (see fig. 7.4.1c). The spermatheca is
commonly surrounded by fat globules.
Male reproductive organs

Male Prostephanus truncatus have a pair of testes, each divided up into five
follicles. The male reproductive system in LGB also contains four pairs of lobes of varying
structure, all converging at the ejaculatory duct. These include the seminal vesicles and
accessory glands (see Scholz, 1997). The sclerotised intromittent organ is continuous with
the exoskeleton and can be manually everted by applying pressure to the abdomen. The
111
Fig. 7.4.1a: Diagram to show
plan view o f fem ale reproductive
tract including the orientation
o f the spermatophore (shaded Ovarioles
in grey) im m ediately after
copulation.
Fig. 7.4.1b: Photograph o f Germarium

LGB oviducts showing m uscle Developing egg
tissue (light m icroscope low Mature egg
power). Ovary base
O-ovary base, LO-lateral Lateral oviduct
oviduct, B-bursa.
Blind-ending-duct
Fig. 7.4. lc: Photograph o f Sperm reservoir
LGB spermatheca (light m icro Accessory gland
scope, low power). Bursa
SR-sperm reservoir, BED- Spermatophore
blind-ending-duct, AG-
accessory gland, B-bursa,
F-fat globules.
Fig. 7.4. la.
i------------- 1 i---------- 1
0 .1mm 0 .1 m m
Fig. 7.4.1b. Fig. 7.4.1c.

Fig. 7.4. Id: Photograph o f
male LGB intromittent organ
(light m icroscope, low power).
A-aedeagus, P-paramere.
I--------------------- 1
0.3mm
i-------------------- 1
0.1mm
Fig.7.4. l e (left): Photograph o f LGB

spermatophore (light m icroscope, low power).
NB: V-shaped invagination at its base.
Fig.7.4. If (above): Close up photograph o f the

tip o f the spermatophore (light m icroscope,
medium power), showing fine tube inside.
0 .1 m m
three lobes splay out of a cylindrical sheath that surrounds them when they are contained
within the body (see fig. 7.4.Id). The central lobe (aedeagus) is roughly cylindrical in
shape with a central duct running through it. The two lateral lobes (parameres) are also
cylindrical up until their ends where they are tapered and possess five hairs. The lateral
lobes tend to be slightly longer than the central shaft.
Male LGB deliver sperm in the form of a spermatophore see fig.7.4. le. This
oval shaped package of sperm has a roughly triangular shaped infolding part, which can
evert. A very long thin tube is also folded inside the spermatophore (see fig.7.4.If). The
spermatophore is delivered into the bursa copulatrix of the female during copulation. The
spermatophore is orientated such that the inverted parts are opposite the duct leading to the
spermatheca (see fig. 7.4.1a). It is difficult to determine if there is any matrix delivered in
the ejaculate. On some occasions the sperm seemed to be contained within a matrix,
however, dispersing the sperm causes this to disappear leading to the conclusion that a high
density clump of sperm may have physical properties like a matrix.
7.4.2 Sterilization of males for calculation of P2

No adults irradiated at the two highest doses (60GY and 120GY) survived
beyond two weeks after the dose was administered. All other treatments had approximately
50% survivorship up until this time. No further trend in survivorship with dose was
apparent from these trials. Viable offspring (those developing at least to the larval stage),
were produced from virgin females mated to males irradiated with doses under 15GY. Only
1 dead egg (no viable offspring) was produced from all matings with males irradiated with
15GY or over. Dissection of a male irradiated with 15GY revealed healthy looking, active
sperm in the testes.
Since survivorship was low and females mated with sterilized males did not lay
eggs this work was abandoned.
7.4.3 Determination of sperm number per ejaculate for two

consecutive matings of males kept in different sociosexual
environments.
Average measures of male mating investment

Time before mating
The time taken for a male to mate was constrained by experimental design to be
between 0 and 25 minutes. Ninety two out of 126 mating trials resulted in a mating before
25 minutes (73%). The range recorded was from 2-24 minutes and the median (calculated
including individuals that did not mate within 25 minutes) was 8.5 minutes.
Weight changes in males and females
114
Chapter 7; Sperm competition
Both males and females lost weight in replicates where no mating occurred:
males lost a mean of 0.033mg (SE = 0.0051, N=12); females lost a mean of 0.039mg (SE
= 0.0075, N=12). In trials where copulation occured males tended to lose more weight
relative to their unmated counterparts and females tended to gain weight relative to their
unmated counterparts.
Males that mated in their first trial gained a mean of 0.044mg (SE = 0.004mg,
N=15) during the five hour period between mating trials.
Spermatophore size
Of 90 matings, 54 intact spermatophores were recovered, 19 burst
spermatophore were found and in 17 cases no spermatophore was found. In two cases
mistakes were made in the dissection and ejaculate data from these replicates have been
excluded from all further analysis. In no cases were more than one spermatophore per
mating recovered. Of the 54 intact spermatophores the mean width was 0.23mm (SD ±
0.064mm) and the mean length was 0.59mm (SD ± 0.173mm).
Sperm number per ejaculate

From all ejaculates sampled the mean estimate of sperm number per ejaculate
was 35600 (SD=29700, N=90). Sperm number was not found to be significantly
correlated to spermatophore size. The mean standard error of sperm number, incurred from
taking sub-samples from the whole was 15.8% (N=79) (each error expressed as a
percentage of its mean).
Comparison of male mating investment between 1st and 2nd matings:

Are males sperm limited?
Proportion mated
A larger proportion of mating trials resulted in a mating for second mated males
in comparison to first-mating males (76.2% compared to 71.4%). A direct comparison of
these two proportions is complicated by the fact that all those males taking part in the
second mating trials have already proved capable and ready to mate since they have showed
a 100% rate of mating in the first trials. Therefore the information lacking to interpret these
data is whether the mating performance in the first trial is a consistent feature of individuals
ie. non-maters will never mate and maters will always mate. If this was true then beetles
became less likely to mate in a trial if they have recently been mated (76.2% success
compared to 100% success), however, if there is no consistency for each individual
between trials then beetles became more likely to mate during a second mating trial (76.2%
success compared to 71.4% success).
Time before mating

A frequency histogram of time before mating for first and second matings
pooled for all treatments is shown in fig. 7.4.3g. Time before mating was numerically
115
lower for first matings compared to second matings, but this difference is not significant in
this sample: median for first mating 7.5 minutes (N=84); median for second mating 9.5
minutes (N=42); W ilcoxon signed-ranks test for replicates where the m ale mated twice,
z = 1 .0 7 , N = 3 2 , p= > 0.1.
calst mating
■2nd mating
7 9 11 13 15 17 19 2 1 23
Time before mating (minutes)
Fig. 7.4.3g: Distribution of time elapsed before mating for first and second matings. First matings, N=60;
second matings, N=32. Wilcoxon signed-ranks test for replicates where the male mated twice, z=1.07,
N=32, p=>0.1.
W eight changes in m ales and fem ales

In 1st mating trials where the pair copulated, m ales lost a mean o f 0.052mg
more than their unmated counterparts and females gained 0.052m g more than their unmated
counterparts (N =15). In 2nd mating trials where the pair copulated, males lost a mean o f
0.019m g and fem ales gained a mean o f 0.028m g more than their unmated counterparts
(N = l 1). A two tailed t-test comparing male weight loss for first mated males and second
mated m ales found that m ales lost significantly less weight during second matings (t=-2.89,
df=23.5, p<0.01). Likew ise, fem ales gained significantly less weight (in fact there was a
net loss o f weight) when mated with m ales who had been mated previously with another
fem ale compared to when mated with first mating m ales (t=-3.26, d.f.=23.5, p<0.01).
Sperm atophore size

Spermatophores produced during the 2nd matings were significantly smaller
than those produced during the 1st matings: t-test on widths o f 1st vs. 2nd m atings p=0.0012
(2 tailed), t-test on lengths o f 1st vs. 2nd matings p=0.025 (2 tailed) (see fig. 7.4.3h).
116
4
N=41 N =13
0.62m m
± 0 .0 3 mm 0.50m m
± 0.05m m
< > < »

0.24m m 0.18m m
± 0 .009m m ± 0 .0 15mm
1st M ATING 2nd M ATING
Fig. 7.4.3h. Diagram to show the size difference between spermatophores delivered in first and second
matings.
Sperm num ber p e r ejaculate

Considering only replicates where males mated twice, the mean number o f
sperm delivered in second ejaculates is less than half that delivered in first ejaculates (see
fig. 7.4.3i). Significantly more sperm per ejaculate are delivered in the first mating
com pared with the second (W ilcoxon signed-ranks test: z= -3.693, N = 32, p<0.001).
60000 -
50000 H
3
£
a? 40000 -
& 1»:
30000
“ 20000 i
C/5
C
s
2 10000 i
1st mating 2nd mating
Fig. 7.4.3i: Mean sperm number per ejaculate for first and second matings (only replicates where males
mated twice). N=32 for each column.
117
Comparison of male mating investment between treatments: Do

males vary their mating investment in response to sociosexual
environm ent?
Proportion m ated
N o treatment effect was demonstrated for the percentage o f replicates that
mated, either for first or second matings (see fig. 7.4.3j).
□ 1st mating (N=28)

o2nd mating (N=14)
Single, fresh Single, Crowded,

conditioned fresh
Fig. 7.4.3j: Bar chart to show the percentage of replicates that mated for all three treatments and for first and
second matings. N=28 for 1st matings and N=14 for 2nd matings for each treatment.
Time before m ating

The median time taken to mate in the first mating (when considering the total
Single fresh median =
population o f all treatments including males who did not mate) is:
6.5 (N = 28), Single conditioned median = 9.5 (N = 28), Crowded fresh median = 6.5
(N =28). A Kruskal-Wallis analysis on the data available indicates that there is a treatment
effect on time before mating: test statistic = 7.43, p=0.0024. This must be interpreted with
caution since the data is incomplete (some males did not mate within the observation time).
I have included this analysis because no significant treatment effect on proportion o f the
sample that mated was found, so the exclusion o f unmated replicates may not have a
significant effect on the overall result.
Sperm atophore size

Only 13 intact spermatophores were recovered from second matings across all
treatments so these were not used for analysis o f treatment effect. The size and shape o f
spermatophores produced by males from each o f the treatments is shown in fig. 7.4.3k. N o
118
treatment effect on shape is apparent. A size index (width x width x length), was used to
compare overall spermatophore sizes between treatments. Considering data from first
matings only,Crowded fresh m ales produced the largest spermatophores (mean size
index = 0.047± S E 0.008, N = 15), and Single fresh m ales produced the sm allest
spermatophores (mean size index = 0.036±SE 0.006, N =16). The differences betw een
spermatophore sizes for each treatment for 1st matings in this sample are not significant:
single factor A N O V A in E xcel 5, F=0.8, p=0.47.
900 i
800 - 0 Single, fresh

°° ©Single, conditioned
700 - • 80 • Crowded, fresh
600 -
0
500 - o*6
o o
400 - o
DO •o
C 300 '
•
200 -
1 00 -
100 200 30 0 400

Width (micrometers)
Fig. 7.4.3k: Scatter graph to show spermatophore length vs. width for first matings only (points separated
by male treatment). Single, fresh, N=16; single conditioned, N=10; crowded, fresh, N=15.
Sperm number p e r ejaculate

N o treatment effect has been demonstrated for sperm number per ejaculate. The
raw data are shown graphically in fig.7.4.31. There is a tendency for m ales kept in isolation
on clean m aize (single, fresh) to deliver relatively small first ejaculates. Sperm number in
the first ejaculate is not a clear predictor o f sperm number in the second ejaculate.
119
160000
SINGLE, FRESH
! 140000
^ 120000
fc 1 0 0 0 0 0
cu
S-H
80000
i§ 60000
g 40000
& 20000
co
160000
SINGLE,
140000 CONDITIONED
120000
100000
80000
60000
40000 1
20000
160000 i CROWDED, FRESH

J 140000
£ 120000
"a?
^ 100000
<D 8 0 0 0 0
3 60000
e
40000
<D
£ 20000
0 "
1st mating 2nd mating
Fig. 7.4.31: Three charts to show sperm number estimates for ejaculates of first and second matings of
males in three treatments: single, fresh; single, conditioned; crowded, fresh. Dotted lines indicate replicates
where the male did not mate with the second female within the time given.
120
Means and standard deviations of sperm number across treatments and

associated statistical analysis are given below in table 7.4.3m. The distribution of sperm
number per ejaculate for first matings was transformed by square-rooting all values to make
the data approximate to a normal distribution. A one way ANOVA (Excel 5) of the
transformed data showed that there was no significant difference between the treatments:
F=1.46, p=0.24).
Table 7.4.3m: The mean number of sperm per ejaculate delivered by males exposed to three different
sociosexual environments on their first mating.
1 1 Mean sperm number 1 SD..............[ N

| [ per ejaculate | [
I Single, fresh | 32400 j 27800 I 21 j
| Single, conditioned | 47600 j 28600 | 17 '$
I Crowded, fresh 1 51000 | 40900 | 20

i
The mean number of sperm per ejaculate delivered during second matings for
the three treatments are shown in table 7.4.3n. No treatment difference for second matings
is suggested by the data. No statistical analysis for a difference was conducted to confirm
this because of the low sample sizes.
Table 7.4.3n: The mean number of sperm per ejaculate delivered by males exposed to three different
sociosexual environments on their second mating.
I
I
j
^ Mean sperm number per I
| N
ejaculate | 1|
1 1
| Single, fresh | 18000 | 8 1
" .......... f r jull juu .u.,1 i in i.ic r I. „ _uluu. jliiiji.iji i 1.1. . I . . ijiijijii | .IT,
1
| Single, conditioned 13500 | 8 1
1
1 1: I
I Crowded, fresh 1 15900 1 5
1 \ „ , . r...... 1..............
Controls
Time o f day control
Males allowed to mate for the first time in the afternoon gave a mean sperm
number per ejaculate that fell between the values obtained for first mating males in the
morning and second mating males in the afternoon (see fig. 7.4.3o). The data were
normalised by square rooting all values. Comparisons between categories were tested for
significance using a t-test (Excel 5), and no significant differences were demonstrated (1st
mating a.m. vs. 1st mating p.m., p=0.14; 2nd mating p.m. vs. 1st mating p.m., p=0.30).
The standard deviation bars in fig. 7.4.3o illustrate that the 1st mating (pm) sample has a
121
standard deviation more comparable to that o f the 1st mating (am) sample than the 2nd
mating (pm) sample.
Residual sperm num ber p e r fem ale

The range of sperm number estim ates for five fem ales not used in mating trials
was 2 50 - 3500 sperm, mean = 1775 sperm (see fig . 7.4.3o).
70000
N = 34
0 6 0 0 0 0 -j
a
1 !
•jj1 5 0 0 0 0 H N = 15
±
£ 40000 i
D
C.
»3
® 30000 ]
N = 21
g 2 0000 -i
i
o
^ 10000^
N= 5
1st 2nd 1st Residual

mating mating mating sperm
(am) (pm) (pm) estimate
Fig. 7.4.3o: The mean sperm number estimates for time of day control and estimate of residual sperm in
females are compared to previously obtained standards. Error bars above the bar are standard deviations and
on the bar are standard errors.
7.5 DISCUSSION
7.5.1
Female reproductive organs
The number of ovarioles is very variable between Families o f beetles,
Scarabaeinae have just one yet 200 are found in M eloe (Cucujidae) (Wigglesworth 1972).
The five per ovary found in LGB is a relatively low number. This is to be predicted by the
known life-history characters o f this species. LGB is a relatively long-lived species with a
long laying period that starts about fiv e days after eclosion and continues until death, 10-24
weeks later (Li, 1988). This gives rise to an average lifetim e fecundity o f 320 eggs per
female (Li, 1988). More ovarioles m ay be expected in species that have a short burst o f
reproduction and require many eg g s to be mature at once. Li found the size o f clutch to vary
122
Chapter 7; Sperm competition
between 3 and 14 eggs per blind-ending tunnel in maize. It is perhaps surprising that the
upper limit of clutch size is not 10, i.e. the number of ovarioles.
The bursa is of comparable size to one egg or one spermatophore. It is

muscular in nature and violent contractions have been observed following mating and
dissection of the reproductive tract. It is unknown whether the contractions are an artifact of
the dissection procedure, though similar contractions have been noted in other studies and
are thought to be initiated by a component of the male ejaculate (see chapter 6 of Eberhard,
1996). Sperm may be moved around the female tract during these contractions. If different
males initiate different levels/patterns of contraction that ultimately affect the chances of
their sperm fertilising the female’s eggs, then these contractions could constitute one form
of cryptic female choice.
The presence of a female sperm-storage organ in this species means that inter
male competition at the level of the gamete is likely to be common. Storage of sperm means
that ejaculates inseminated relatively far apart in time may be maintained in a viable state in
the female. Fat bodies found around the spermatheca may function as a reservoir of food to
be supplied via the accessory gland to the stored sperm, thereby prolonging their life and
reducing the number of times a female needs to mate to ensure a reliable source of viable
sperm. Eberhard (1996) highlights that other parts of the female reproductive tract could be
used as more temporary sperm-storage structures, for example the bursa and/or the
oviducts. The bursa is sometimes a hostile environment to sperm (Eady, 1994). Sperm
reaching these areas may, however, gain easier access to eggs as they are laid. Therefore,
in situations of high population density and relatively high incidence of mating,
spermathecal sperm may not play a major role in the ‘mating game’. However, if females
refrain from mating, perhaps during dispersal or because they are effectively isolated in a
tunnel, then spermathecal sperm, which is of potentially extended lifespan from nutrients
supplied by the spermathecal gland, may play a larger role.
Villavaso (1975), quoted in Eady (1994), has shown that contractions of

muscular tissue in the spermatheca in the boll weevil Anthonomus grandis expel sperm
from the spermatheca. Eady (1994) envisages that a similar process may occur in
Callosobruchus maculatus. In C. maculatus the tip of the comma-shaped sperm-storage
chamber may be pulled round by the contraction of such a muscle stimulated by the
straightening of the comma as it fills with sperm. The function of the blind-ending duct in
LGB is unknown. If there is a physical connection between the blind-ending duct and the
receptum seminalis then perhaps sperm from previous matings can be flushed out down
this duct. In some species the duct leading to the spermatheca is vastly elongated and often
convoluted in form (see examples cited in chapter 7 of Eberhard, 1996). Eberhard proposes
that these structures have evolved since they allow cryptic female choice for ejaculates that
can overcome such a barrier. In contrast the duct leading from the spermatheca to the point
123
of fertilization is often much shorter (Eberhard, 1996) so presumably the female is better
able to manipulate fertilisation prior to storage. No such convolutions or elongation of ducts
leading to or from the spermatheca were found in LGB.
Male reproductive organs

LGB transfer sperm using a spermatophore. It is not known exactly how
sperm are transferred from this spermatophore to the spermatheca, but it is likely that both
the long thin tube folded inside the spermatophore and the inverted region of the
spermatophore are everted for this purpose. It seems unlikely that the intromittent organ in
LGB can be used to scoop out previous male’s ejaculates from the spermatheca since it is
too short and wide. The hairs at the end of the parameres could, however, serve this
function for any sperm/spermatophore material that may remain in the bursa. Removal of
material may be particularly important if the remains of previous spermatophores act as a
barrier to later ejaculates. The length of the parameres is approximately half that of the
bursa, but it is feasible that the intromittent organ is articulated deeper into the female by
structures at its base. It is not known how the spermatophore is delivered from the male to
the female in LGB. It is likely that the spermatophore is at least partly formed within the
female tract and in some species spermatophore shape is determined by the shape of the
female bursa (Gerber, 1970). I propose that this is not the case in LGB since although the
shape of the spermatophore is similar to that of the bursa, two normal shaped
spermatophores have been dissected from the bursa of a female allowed to mate with two
males in quick succession (pers. obs.).
LGB deliver an oversized ejaculate in the sense that many more sperm are
delivered than can feasibly be stored in the female’s spermatheca. A popular theory to
explain oversized ejaculates is that they are more effective at diluting down competitor’s
sperm either where ejaculates are able to mix in a so called ‘raffle contest’, or if previously
inseminated sperm can be displaced out of a sperm storage organ. This idea is supported by
data collected by Eady (1995) in his study of Callosobruchus maculatus. In this study male
ejaculates were manipulated by allowing males prior matings. Males delivering smaller
ejaculates were found to fertilise a smaller portion of offspring when they were the second
male to mate in twice-mated females. Cases where males transfer more sperm when
competition between ejaculates is likely to be higher also support this idea that increasing
sperm number can increase paternity in multiply-mated females.
Paternity o f offspring from doubly mated females:

Technical difficulties prevented direct determination of whether males can
secure paternity of eggs laid by females who have already mated. The indirect evidence that
this is the case is strong. A review of all insect studies currently available shows that the
proportion of offspring sired by the second male to mate in a doubly mated female (P2
value) is variable and often such that the last male fathers more than half the offspring (last-
124
male sperm precedence). The fact that LGB males invest highly in ejaculates and spend
much time associated with non-virgin females also hints that total first-male sperm
precedence is unlikely. Indeed, one mating is not enough to supply enough sperm for a
female’s entire reproductive lifetime (assuming females do not die for other reasons).
Mating frequency of beetles observed within plant hosts in this study (chapter 6) in LGB
has been found to be higher than that required for fertilisation alone. Ridley (1989) found
that high mating frequency is associated with high P2values. One measure of P2, although
interesting, is not really enough to quantify the relative costs and benefits of multiple
mating, mate guarding and investment in ejaculate. P2is determined by many factors such
as time between matings, previous mating history, availability of a suitable host as well as
individual differences between males and females and interactions between these factors.
Indeed, increasing the number of male mates from two to three can greatly complicate
patterns of paternity (Zeh and Zeh, 1994). Efficient methods of tracing paternity (for
instance using molecular markers) will help scientists to obtain the kind of direct
measurements of reproductive success needed to test hypotheses about mating tactics.
Ejaculate investment:
It was possible for an estimate of sperm number per ejaculate to be determined
for LGB. The mean standard error associated with sampling was fairly large (15.8% of the
mean). Possible contamination of ejaculate samples with sperm already stored in the female
spermatheca was a maximum of about 10% of total recorded, though it must be stressed
that stored sperm tended to remain in the sperm reservoir whilst the spermatophore was
removed from the female tract, so this error is likely to be much less than the maximum. It
is impossible to quantify the error associated with imperfect removal of the full ejaculate
from the female although containment of sperm within spermatophores made full removal
easier. Of course these errors were the same on average for all treatments. The large
variation in sperm number per ejaculate for all treatments makes the errors associated with
the methods less crucial. That this variation is not an artifact of the methods used is
demonstrated by the consistent difference detected between first and second ejaculates.
This study did not find any change in some measures of ejaculate investment
for males placed in three different sociosexual environments (single male on fresh grain,
single male on conditioned grain and crowded males on fresh grain). Ejaculate investment
was, however found to be extremely variable (standard deviation of sperm number in first
ejaculates was, on average 76% of the mean). Neither male size nor female size were found
to account for this variation. Male age was not strictly controlled in these experiments and
could be a main determinant of ejaculate size. Fox et al., (1995) propose that male body
size does often correlate with ejaculate size, but that this relationship is only detectable
when age and other factors are controlled for. Female size has also been correlated with
ejaculate size in the cricket, Acheta domesticus (Gage and Barnard, 1996). Larger females
125
are more fecund and males could therefore preferentially invest more ejaculate resources in
these females if such resources are limited.
Male LGB are sperm limited. Both sperm number, spermatophore size and
male-weight loss were found to decrease for the second matings compared with the first.
They delivered smaller ejaculates if allowed to mate twice within five hours instead of being
more reluctant to mate. Unfortunately, it was not conclusively demonstrated that these
smaller ejaculates were a feature of second matings as opposed to ejaculates delivered in the
afternoon. Males lost a detectable proportion of their body weight (an average of 1.6% of
body weight) after the delivery of ejaculates, which is less than the 5% body weight loss
found by Fox et al., (1995) in another stored product beetle, Callosobruchus maculatus.
There is not enough information currently known about other species similar to
LGB for a full appraisal of the significance of the results obtained in this study. The results,
however, suggest that the large numbers of sperm delivered per ejaculate are an adaptation
to maximise paternity in the face of sperm competition in this polyandrous species.
Ejaculate size is extremely variable in this species and alternative mating tactics could
possibly maintain this variation. Female Factor and male crowding are apparently not used
as cues by males to alter their investment in ejaculates.
126
Chapter 8: Discussion
CHAPTER 8: DISCUSSION
8.1 Evolution of the aggregation pheromone signal/

response:
Sex-specific ornaments could arise from a variety of non-exclusive
mechanisms of which sexual selection is just one possibility. Current ideas are summarised
by Andersson (1994) and are given below. These ideas can be used to address the
question, ‘why is it only LGB males who produce aggregation pheromone?’. Each of
these possible mechanisms will be discussed in reference to the system found in LGB in
order to review the evidence to date. Although only mate acquisition benefits of signalling
are tested in this thesis, possible collective feeding benefits derived from the initiation of
aggregations will also be discussed.
A. Male ornaments may have evolved because of:
1. Pleiotropic gene effects

2. Selection of ecological sex differences
3. Males being unprofitable prey for predators
4. Male contests
5. Female choice and mating preferences
B. Female preferences for male ornaments may have evolved because of:
1. Fisherian self-reinforcing selection

2. Indicator mechanisms
3. Selection for species recognition
4. Direct phenotypic benefits to choosy females
5. Selection of the sensory system in other contexts (sensory bias)
6. Advantages in the timing of reproduction (mating synchronization)
Taken from chapter 1 in Andersson (1994).
A1 Pleiotropic gene effects: It is possible that male ornaments could be

by-products of other sex-linked genes that serve a different function. In this case the
pheromone signal could be a biochemical waste product of, for instance, sperm production.
Alternatively the genes for pheromone production could be linked to those for sperm
production and could be expressed since those for sperm production are maintained in the
population through natural/sexual selection. Aggregation pheromone signalling in LGB
127
does however, ‘enhance mating and reproductive success’, and it is therefore unlikely that
this trait is a selectively neutral by-product of pleiotropy.
A2 Selection of ecological sex differences: Andersson proposes

feeding differences as an example of ecological sex differences. I take the phrase,
‘ecological sex differences’ to encompass any sex-specific difference in traits favoured by
natural selection other than those that are sexually selected. Such ecological sex differences
need to be identified if the primary benefit of aggregation pheromone signalling is host
conditioning, since otherwise, why do females never signal?
If the main benefit of signalling is host conditioning then we would predict

females to signal until the host is suitable as a site for oviposition. Perhaps female
Prostephanus truncatus do signal when they arrive on hosts that require conditioning and
we have not detected such a signal because so far we have looked for signalling on stored
products which may not require conditioning. Its a long shot, and there is no evidence for
this. We may still need to infer a mate-acquisition benefit for males to explain why they
signal on stored products.
Males and females may differ in their propensity to disperse. Perhaps females
are never the first to arrive at a new host. Sex differences in the costs of dispersing, for
instance the ability to fly, or ability to store food for the journey, may have created a
dichotomy. The benefits of dispersing may also be different between the sexes. The
evidence we have of current populations is that there is no sex difference in flight duration
or propensity to fly (Fadamiro, 1995). Indeed, a female bias has been found in the
dispersing population (Scholz, 1997 and chapter 5 this thesis). Therefore the evidence to
date does not support this hypothesis.
The key point may be that eggs and larvae are more vulnerable to predation
than adults. Natural selection may tend to reduce the use of signalling in the presence of
one’s offspring. Therefore ovipositing females should not signal, and this might also
explain why males decrease their signal in the presence of females.
A3 Males being unprofitable prey for predators: Baker and Parker

(1979) proposed that male signals (in this case in the form of conspicuous plumage in
birds) may be a signal to predators. The ‘unprofitable prey hypothesis’ proposes that
certain individuals of a population (perhaps males) are better able to escape if predators
pursue them or are perhaps smaller. The predator will be more efficient if it ignores these
individuals and only pursues the prey items that provide more resources per effort spent
catching them. Predators may leam/evolve to avoid unprofitable prey if they can be
distinguished from profitable prey by a reliable signal.
128
I propose that the costs of being conspicuous and therefore attracting a greater
number of predators far outweigh any possible deflection of predation onto other
conspecifics in LGB. LGB adults tend to decrease their signalling effort as an aggregation
grows in direct opposition to the predictions of this theory (Prof.D.R.Hall pers. com.).
A4 Male contests: It is proposed that male signals such as plumage

characters and song in birds may serve to ward off male competitors (see Fisher, 1930
referenced in Andersson, 1994). However, Prostephanus truncatus males ‘prefer’ to
follow the same signals as females and therefore ‘strong’ signals do not result in avoidance
of other males. Males may tunnel a short distance away from male competitors and there is
the possibility that males will space themselves further away from ‘strong’ signals perhaps
after using the strongest signal to locate higher quality hosts (see effect of host on signal
discussion in chapter 3). Male avoidance of strong signals has so far only been
demonstrated for the signal produced by the equivalent of many males signalling in the
same patch of food (Schlyter et al., 1987).
A5 Female choice and mating preferences: Sex-specific signalling

could have arisen and be maintained if females preferentially mate with signalling males.
This is the hypothesis that I have been testing in this thesis. Previous evidence showed that
the aggregation pheromone signal is a determinant of conspecific distribution which, in
turn, is obviously a crucial determinant of mating opportunity. Work presented in this
thesis shows that variation in the aggregation pheromone signal is perceived by responding
beetles and does determine the relative numbers of beetles arriving at each signaller. It has
been shown that this is where female choice between signals ends in LGB, and other cues
take over as determinants of mating success and probably fertilization success.
B1 Fisherian self-reinforcing selection: Fisherian self-reinforcing

selection is a mechanism proposed to explain female preferences for male traits that do not
necessarily confer fitness benefits directly, or indeed indicate good quality. It is proposed
that male traits can be selected for because they confer a mating advantage because females
preferentially choose such males. Females choosing such traits benefit by producing ‘sexy
sons’. Thus the alleles for the female preference of the male trait favoured by Fisherian self
reinforcing selection will be genetically linked to the alleles that produce the trait. If the
preference is for the most extreme form of a trait, for instance the longest tail, then the trait
may become exaggerated in what is called a runaway process, until it confers enough of a
disadvantage to the bearer to be arrested by natural selection. This hypothesis has not been
tested here.
B2 Indicator mechanisms: Honest indicating of general quality is likely if

production of the signal is energetically costly and therefore disproportionately so to poorer
quality males. There is some evidence that this is the case in LGB (see discussion in chapter
129
3). Favoured aggregation-pheromone signals in LGB may be both an indicator of a

physiologically healthy male and also one who is able to locate a good host. If variation in
these aspects of male quality are partly heritable then females would benefit from acquiring
these ‘good genes’ for their offspring.
B3 Selection for species recognition: Species recognition through sex

and aggregation-pheromone signals in general is well supported since many of these signals
are species specific. Even the blend of chemical components and enantiomers have been
shown to be species specific, and have been inferred to be an isolating mechanism between
some sympatric species of bark beetles (Wood, 1982). The pheromone from another
Bostrichid beetle, Rhyzopertha dominica is attractive to LGB (Hodges et al., (1983).
However, there is no evidence to date that the reverse is true (Rick Hodges pers. comm.)
and only LGB and its predator Teretriosoma nigrescens are consistently caught in traps
baited with LGB aggregation pheromone (T1 and T2). Aggregation pheromone in LGB
may be related to species recognition, though there is no evidence that this is a primary
function.
B4 Direct phenotypic benefits to choosy females: Direct phenotypic

benefits to choosy females could include the selection of superior hosts if host quality
influences the pheromone signal (see effect of host type on signal in LGB, chapter 3). This
is broadly analogous to the correlation between female attraction through male bird song
and food abundance in the territory found in red-winged Blackbirds (Ewald and Roher
1982, quoted in Andersson, 1994). Increasing the food available to male blackbirds led to
them attracting more females to their territory. Females may also benefit from selecting a
‘good phenotype’ (which need not be heritable) in males since LGB males possibly assist
during oviposition, by reducing the entry of predators into the tunnel system, and by
constructing the main tunnel into the host. Females may assess these particular benefits
more efficiently by the pushing behaviour observed prior to mating (see chapter 6). Data
presented here are consistent with, but do not test this hypothesis.
BS Selection of the sensory system in other contexts (sensory

bias): Female response to male stimulants may have evolved in another context.
Aggregation pheromones may be a mimic of plant volatiles that conspecifics follow to
locate a host (‘sensory trap’ tactic). An alternative view is that volatiles released by beetle
feeding were exploited by other beetles for host location. However, LGB is not known for
its specific reactions to host volatiles (Pike et al, 1994), and the chemical structure of
aggregation pheromone in LGB is not particularly characteristic of that of plant volatiles
(David Hall pers. comm.), so this hypothesis is not supported by the data available at the
moment.
130
B6 Advantages in the timing of reproduction (mating

synchronisation): Andersson points out that, ‘sex ornaments and courtship may
function after pair formation and are not involved in competition over mates’. Sex-specific
ornaments may initiate appropriate behaviour/physiological changes in readiness for
copulation and fertilization. This does not appear to be the case in LGB since manipulation
of the pheromone signal did not detectably alter either precopulatory courtship behaviour or
ability to transfer sperm. The possibility that aggregation pheromone might determine some
aspects of cryptic female choice have not been tested. Processes such as sperm uptake to
the spermatheca could conceivably require the pheromone signal as a stimulant, but it is
more likely that any chemical stimulants required are contained within the ejaculate (see
Eberhard, 1996).
8.2 Alternative mating tactics:

Features of mate acquisition/ fertilization however, tend to be very plastic
within individuals and males may switch strategies in response to features of the plant host,
conspecific signals and behaviours, and male physiological state. The possibility that males
may be polymorphic in respect to mate acquisition behaviour has not been indicated by the
results obtained in this study. Although variation in key traits (such as signalling, body
size, sperm investment, pushing behaviour) that may influence reproductive success in
males have been shown to be extensive in LGB, discontinuous variation was not found
when males were placed in the same conditions.
If we assume that LGB males are behaving optimally then we can infer that
maximum reproductive success of males should gained by: locating a suitable host,
constructing a tunnel system into which males can lure females; mating the female many
times a day with oversized ejaculates; and guarding her during oviposition. By-passing the
construction of a tunnel system altogether and permanently specialising in acquiring mates
as a roving male does not appear to have evolved in LGB. The absence of a permanendy
roving male strategy might be a consequence of the difficulty in gaining access to a paired
female once she has entered another male’s tunnel system. Males will mate with females
that they encounter away from a tunnel system. However, if an uncolonised suitable host is
available, males will always start to construct a tunnel system, even though this effort is far
beyond that required for feeding alone. Males will follow other male’s signals when
dispersing themselves, presumably as an efficient way of locating a high quality host.
Observations of beetles arriving at a large host would show whether males and females
differ in the precision to which they locate the signaller and the degree to which they stop
searching once contact is made on the host. I would predict that females are more likely to
search over the surface of a host for the male signaller and his tunnel system than males,
who may actively space themselves a short distance from competing males.
131
8.3 The limitations of an evolutionary approach:

The discussion above is riddled with words like ‘maybe, possible and likely’
which are indicative of the inherently low predictive power of an evolutionary approach.
Few, definite conclusions can be made regarding the evolution of male-produced
aggregation pheromone in LGB, but the process of evaluating the possibilities does
highlight important gaps in our knowledge of this pest. An appreciation of the benefits and
costs of pest behaviours is crucial for the prediction of the effect that any man-made
manipulation of such behaviours may have on the population dynamics of the pest.
8.4 Practical applications of this work:

This study has for the first time developed a bioassay for detecting differences
in the aggregation-pheromone signalling of individual beetles. Use of single beetles has not
only helped to understand the evolutionary function of aggregation pheromone in LGB, but
has also shed some light on its potential use in pest control.
Monitoring
Single insects have been shown to be an effective bait for LGB when loaded
into very simple funnel traps. This paves the way for a host of studies that could investigate
natural signalling and response in LGB. Live LGB adult males could also be used in the
construction of a very low cost trap when artificial pheromone vials are unavailable or
inappropriate. Fears of increasing the infestation rate of stores by placing traps near them
could be reduced if single-male-baited trap are used instead of artificial pheromone lures in
cases where the trap cannot be placed further from a food store.
Courtship disruption
The aggregation pheromone signal is not used by Prostephanus truncatus to
elicit courtship behaviour on contact and therefore the chance that mating behaviour can be
disrupted at this stage by interfering with this signal is minimal.
Mass trapping/ manipulation o f dispersal patterns

Beetle perception and choice between the natural variation in aggregation
pheromone signal is an important consideration for any proposed programme of mass
trapping or mating disruption (through influence on dispersal) where artificial sources are
presented as point sources. The greater the extent to which beetles invest in choosing
between signals, the more efficient artificial trapping could be, provided the artificial lure is
engineered to be more attractive than the natural signal. Care should be taken not simply to
use high dose lures without considering that beetles may land some distance from high
concentrations of pheromone to avoid the high levels of intra specific competition that this
would naturally indicate.
132
LGB have been shown to become habituated to aggregation pheromone.

Saturating stores with pheromone to disorientate LGB would be too risky and more likely
to promote infestation than reduce it. There could be pheromone analogues, however,
which do not elicit the orientation response, but yet lead to desensitisation of insects to the
pheromone. Such analogues could be identified and deployed in stores.
Host selection
There is some evidence that host characteristics can influence LGB’s
aggregation signal (chapter 3). Identification of the factors that limit the male signal could
lead to a method of suppressing the signal for beetles feeding in stores, thereby making
these colonising beetles less attractive than those signalling on other, as yet unidentified
hosts in the environment away from stores.
8.5 Suggestions for future work:

• Identify the spatial distribution of LGB away from stores and identify the
important hosts used by LGB in the natural environment (field based).
• Confirm that host identity influences the attractiveness of aggregation

pheromone signal in LGB (field or laboratory based) and identify which
aspects of the host are important (laboratory based).
• Use hosts baited with single male signallers to investigate the behaviour of
arriving beetles (field based).
• Determine the nature of ejaculates delivered by males during multiple

mating of females within a tunnel system to assess the function of multiple
mating behaviour (promote success in sperm competition and/or provide
nutrients to the female) (laboratory based).
• Establish whether male presence in the entrance of tunnel systems prevents

entry of predators such as Teretriosoma nigrescens (laboratory based).
133
REFERENCES
Agosta, W. C. (1992). Chemical communication, the language of pheromones. Scientific

American library, New York.
Akers, R. P., H. K. Preisler, and D. L. Wood. (1993). Interactions between components

of the aggregation pheromone during chemotaxis by the bark beetle Ips
paraconfusus. Journal o f chemical ecology., 19:863-879.
Alcock, J. (1982). Natural selection and communication amoung bark beetles. Florida
entomologist, 65:17-32.
Alcock, J. (1994). Postinsemination associations between males and females in insects:

The mate guarding hypothesis. Annual review o f entomology, 39:1-21.
Andersson, M. (1994). Sexual selection. Princeton university press, Princeton, New

Jersey.
Baker, R. R., and G. A. Parker. (1979). The evolution of bird coloration. Philosophical
transactions o f the Royal society o f London B, 287:63-130.
Bartelt, R. J., D. G. Carlson, R. S. Vetter, and T. C. Baker. (1993b). Male produced

aggregation pheromone of Carpophilus mutilatus (Coleoptera, Nitidulidae). Journal
o f chemical ecology., 19:107-118.
Bartelt, R. J., and D. G. James. (1994). Aggregation pheromone of Australian sap beetle,
Carpophilus davidsoni (Coleoptera, Nitidulidae). Journal o f chemical ecology,
20:3207-3219.
Bartelt, R. J., K. L. Seaton, and P. F. Dowd. (1993a). Aggregation pheromone of

Carpophilus antiquus (Coleoptera, Nitidulidae) and kairomonal use of C. lugubris
pheromone by C. antiquus. Journal o f chemical ecology., 19:2203-2216.
Bertram, S. L., and T. D. Paine. (1994). Influence of aggregation inhibitors (Verbenone

and Ipsdienol) on landing and attack behaviour of Dendroctonus brevicomis
(Coleoptera, Scolytidae). Journal o f chemical ecology., 20:1617-1629.
Birgersson, G., G. L. Debarr, P. Degroot, M. J. Dalusky, H. D. Pierce, J. H. Borden, H.

Meyer, W. Francke, K. E. Espelie, and C. W. Berisford. (1995). Pheromones in
white pine cone beetle, Conophthorus coniperda (Scwarz) (Coleoptera: Scolytidae).
Journal o f chemical ecology., 21:143-167.
134
Boeye, J., A. Biliwa, H. U. Fischer, J. Helbig, and J. Richter. (1994). The dispersion
pattern of Teretriosoma nigrescens Lewis (Col., Histeridae) after it's release and
monitoring of the occurrence of it's host Prostephanus truncatus (Horn) (Col.,
Bostrichidae) in the natural environment in Togo., 6th International Working
Conference on Stored-product protection., Vol. 2, pp. 1098-1102, Canberra,
Australia.
Borden, J. H. (1967). Factors influencing the response of Ips confusus to male attractant.
Canadian entomologist., 99:1164-1193.
Bossert, W. H., and E. O. Wilson. (1963). The analysis of olfactory communication

among animals. Journal o f theoretical biology., 5:443-469.
Boughton, A., and H. Y. Fadamiro. (1996). Effect of age and sex on the response of
walking Prostephanus truncatus (Horn) (Coleoptera: Bostrichidae) to its male
produced aggregation pheromone. Journal o f stored product research, 32:13-20.
Bowers, W. W., and J. H. Borden. (1990). Evidence for a male-produced aggregation

pheromone in the four-eyed spruce bark beetle, Polygraphus rufipennis (Kirby)
(Col: Scolytidae). Journal of applied entomology., 110:292-299.
Bowers, W. W., G. Gries, J. H. Borden, and H. D. Pierce. (1991). 3-methyl-3-buten-l-

ol an aggregation pheromone of the 4-eyed spruce bark beetle, Polygraphus
rufipennis (Kirby) (Coleoptera: Scolytidae). Journal of chemical ecology, 17:1989-
2002 .
Budenberg, W. J., I. O. Ndiege, and F. W. Karago. (1993). Evidence for volatile male
produced pheromone in banana weevil Cosmopolites sordidus. Journal of chemical
ecology., 19:1905-1916.
Burkholder, W. E., and M. Ma. (1985). Pheromones for monitoring and control of stored-
product insects. Annual review o f entomolology., 30:257-272.
Butler, C. G. (1967). Insect pheromones. Biological review., 42:42-87.
Byers, J. A. (1993). Avoidance of competition by Spruce bark beetles, Ips typographus

and Pityogenes chalcographus. Experientia., 49:272-275.
Byers, J. A., F. Schlyter, G. Birgersson, and W. Francke. (1990). E-myrcenol in Ips

duplicatus - an aggregation pheromone component new for bark beetles.
Experientia., 46:1209-1211.
Cade, W. H. (1984). Genetic variation underlying sexual behaviour and reproduction.

American Zoologist, 24:355-366.
135
Camacho, A. D., H. D. Pierce, and J. H. Borden. (1994). Aggregation pheromones in
Dryocetes affaber (Mann) (Coleoptera: Scolytidae) - stereoisomerism and species
specificity. Journal o f chemical ecology., 20:111-124.
Chambers, J. (1990). Overview on stored product insect pheromones and food attractants.
Journal of the Kansas entomological society, 63:490-499.
Chambers, J., C. P. Morgan, P. R. White, K. Mori, D. E. Finnegan, and D. B. Pinniger.

(1990). Rust red grain beetle, Cryptolestes ferrugineus, and flat grain beetle,
Cryptolestes pusillus - antennal and behavioural responses to synthetic components
of their aggregation pheromones. Journal o f chemical ecology., 16:3353-3372.
Chapman, J. A. (1966). The effect of attack by the Ambrosia beetle, Trypodendron

lineatus (Oliver) on log attractiveness. Canadian entomologist, 98:50-59.
Christy, J. H. (1995). Mimicry, mate choice, and the sensory trap hypothesis. The
American naturalist, 146:171-181.
Crespi, B. J. (1989). Causes of assortative mating in arthropods. Animal behaviour,

38:980-1000.
Cronin, H. (1991). The ant and the peacock. Altruism and sexual selection from Darwin to
today. Cambridge university press, Cambridge.
Cross, M. (1985). Boring into Africa's grain. New Scientist: 10-11.
Darwin, C. (1871). The descent of Man. and selection in relation to sex. Murray. London.
Demianyk, C. J., and R. N. Sinha. (1988). Bioenenergetics of the Larger Grain Borer,
Prostephanus truncatus (Horn) (Col., Bostrichidae), feeding on com. Annals of the
entomological society o f America., 81:449-459.
Dendy, J., P. Dobie, J. A. Saidi, J. Smith, and B. Uronu. (1991). Trials to assess the
effectiveness of new synthetic pheromone mixtures for trapping Prostephanus
truncatus (Horn) (Col., Bostrichidae) in maize stores. Journal o f stored products
research, 27:69-74.
Devlin, D. R., and J. H. Borden. (1994). Efficacy of antiaggregants for the pine engraver,
Ips pini (Coleoptera: Scolytidae). Canadian journal of forest research., 24:2469-
2476.
Dewsbury, D. A. (1982). Ejaculate cost and mate choice. American Naturalist, 119:601-
610.
136
Dickens, J. C., R. F. Billings, and T. L. Payne. (1992). Green leaf volatiles interrupt
aggregation pheromone response in bark beetles infesting southern pines.
Experientia, 48:523-524.
Dickens, J. C., and G. Wigul. (1987). Conspecific effects on pheromone production by

the Boll weevil, Anthonomus grandis Boh. (Coleoptera: Curculionidae). Journal of
applied entomology, 104:318-326.
Dickinson, J. L. (1988). Determinants of paternity in the milkweed leaf beetle. Behavioural

ecology and sociobiology., 23:9-19.
Dowdy, A. K., R. W. Howard, L. M. Seitz, and W. H. Mcgaughey. (1993). Response of

Rhyzopertha dominica (Coleoptera: Bostrichidae) to its aggregation pheromone and
wheat volatiles. Environmental entomology, 22:965-970.
Eady, P. E. (1991). Sperm competition in Callosobruchus maculatus (Coleoptera:

Bruchidae): a comparison of two methods used to estimate paternity. Ecological
entomology, 16:45-53.
Eady, P. E. (1994). Sperm transfer and storage in relation to sperm competition in

Callosobruchus maculatus. Behavioural ecological sociobiology, 35:123-129.
Eady, P. E. (1995). Why do male Callosobruchus maculatus beetles inseminate so many

sperm? Behavioural ecological sociobiology, 36:25-32.
Eberhard, W. G. (1996). Female control: Sexual selection by cryptic female choice.

Princeton university press, Princeton, New Jersey.
Evans, K. A., and J. Bergeron. (1994). Behavioural and electrophysiological response of

cabbage seed weevils (Ceutorhynchus assimilis) to conspecific odor. Journal of
chemical ecology, 20:979-989.
Fadamiro, H. Y. (1995). Flight behaviour and pheromone communication of the Larger

Grain Borer, Prostephanus truncatus (Horn) (Coleoptera: Bostrichidae). Ph.D.
thesis, University of Oxford.
Fadamiro, H. Y., and T. D. Wyatt. (1995). Flight initiation by Prostephanus truncatus in

relation to time of day, temperature, relative humidity and starvation. Entomologia
experimentalis et applicata, 75:273-277.
Faucheux, M. J. (1994). Distribution and abundance of antennal sensilla from 2

populations of the pine engraver beetle, Ips pini (Say) (Coleoptera: Scolytidae).
Annales des sciences naturelles - zoologie et biologie animale., 15:15-31.
137
Faustini, D. L., A. V. Barak, W. E. Burkholder, and J. Leosmartinez. (1990).
Combination type trapping for monitoring stored product insects - a review. Journal
o f the Kansas entomological society., 63:539-547.
Faustini, D. L., W. L. Giese, and J. K. Phillips. (1982). Aggregation pheromone of the

male granary weevil Sitophilus granarius (L). Journal of chemical ecology, 8:679-
687.
Fisher, W. C. (1950). A revision of the North American species of beetles belonging to the
family Bostrichidae. US Department of Agriculture. Misc. publication 698.
Fox, C. W. (1993). Multiple mating, lifetime fecundity and female mortality of the bruchid
beetle Callosobruchus maculatus (Coleoptera: Bruchidae). Functional ecology,
7:203-208.
Fox, C. W., D. L. Hickman, E. L. Raleigh, and T. A. Mousseau. (1995). Paternal

investment in a seed beetle (Coleoptera: Bruchidae) - Influence of male size, age,
and mating history. Armais o f the entomological society o f America, 88:100-103.
Gage, A., and C. Barnard. (1996). Male crickets increase sperm number in relation to
competition and female size. Behavioural ecology and sociobiology., 38:349-353.
Gage, M. J. G. (1991). Risk of sperm competition directly affects ejaculate size in the
Mediterranean fruit fly. Animal behaviour, 42:1036-1037.
Gage, M. J. G. (1995). Continuous variation in reproductive strategy as an adaptive

response to population density in the moth Plodia interpunctella. Proceedings of the
Royal society o f London. B, 261:25-30.
Gage, M. J. G., and P. A. Cook. (1994). Sperm size or numbers? Effects of nutritional
stress upon eupyrene sperm production strategies in the moth Plodia interpunctella
(Lepidoptera: Pyralidae). Functional Ecology, 8:594-599.
Gast, S. J., M. W. Stock, and M. M. Fumiss. (1993). Physiological factors affecting

attraction of Ips pini (Coleoptera: Scolytidae) to host odor or natural male
pheromone in Idaho. Annals o f the entomological society o f America, 86:417-422.
Gerber, G. H. (1970). Evolution of the methods of spermatophore formation in pterygotan

insects. The Canadian entomologist, 102:358-362.
Giblindavis, R. M., T. J. Weissling, A. C. Oehlschlager, and L. M. Gonzalez. (1994).

Field response of Rhynchophorus cruentatus (Coleoptera: Curculionidae) to its
aggregation pheromone and fermenting plant volatiles. Florida entomologist,
77:164-177.
138
Gries, G., R. Gries, A. L. Perez, L. M. Gonzales, H. D. Pierce, A. C. Oehlschlager, M.
Rhainds, M. Zebeyou, and B. Kouame. (1994b). Ethyl propionate - synergistic
kairomone for African palm weevil, Rhynchophorus phoenicis L. (Coleoptera:
Curculionidae). Journal o f chemical ecology, 20:889-897.
Gries, G., R. Gries, A. L. Perez, A. C. Oehlschlager, L. M. Gonzales, H. D. Pierce, M.

Zebeyou, and B. Kouame. (1994a). Aggregation pheromone of the African
Rhinoceros beetle, Oryctes monoceros (Oliver) (Coleoptera: Scarabaeidae). Journal
o f biosciences, 49:363-366.
Guldbrandtsen, B. (1995). Evolutionaere Aspekter af variation i parringssystemer. Ph.D.

thesis, Aarhus Universitet.
He, Y., and T. Miyata. (1997). Variations in sperm number in relation to larval crowding
and spermatophore size in the armyworm, Pseudaletia separata. Ecological
Entomology, 22:41-46.
Helbig, J., and F. A. Schulz. (1994). Studies on the biology of Prostephanus truncatus
(Horn) (Col, Bostrichidae) on wood. Journal o f applied entomology, 117:380-
387.
Heims, D. A., R. A. Haack, and B. D. Ayres. (1991). Variation in semiochemical

mediated prey-predator interaction, Ips pini (Scolytidae) and Thanasimus dubius
(Cleridae). Journal o f chemical ecology., 17:1705-1714.
Hibbard, B. E., and F. X. Webster. (1993). Enantiomeric composition of Grandisol and

grandisal produced by Pissodes strobi and P. nemorensis and their
electroantennogram response to pure enantiomers. Journal of chemical ecology,
19:2129-2141.
Hodges, R. J. (1986). The biology and control of Prostephanus truncatus (Horn)

(Coleoptera: Bostrichidae) - a destructive storage pest with an increasing range.
Journal o f stored products research, 22:1-14.
Hodges, R. J. (1994). Recent advances in the biology and control of Prostephanus

truncatus (Coleoptera: Bostrichidae)., 6th International Working Conference on
Stored-product Protection., Vol. 2, pp. 929-934, Canberra, Australia.
Hodges, R. J., and C. C. Dobson, (in press). Laboratory studies on behavioural

interactions of Prostephanus truncatus (Horn) (Coleoptera: Bostrichidae) with
conspecifics, synthetic pheromone and the predator Teretriosoma nigrescens
(Lewis) (Coleoptera: Histeridae). Journal o f stored products research.
139
Hodges, R. J., D. R. Hall, P. Golob, and J. Meik. (1983). Responses of Prostephanus
truncatus to components of the aggregation pheromone of Rhyzopertha dominica in
the laboratory and field. Entomologia experimentalis et applicata., 34:266-272.
Hodges, R. J., D. R. Hall, J. N. Mbugua, and P. W. Likhayo. (1998). The responses of

Prostephanus truncatus (Coleoptera: Bostrichidae) and Sitophilus zeamais
(Coleoptera: Culculionidae) to pheromone and synthetic maize volatiles as lures in
crevice or flight traps. Bulletin of entomological research, 88.
Hodges, R. J., and V. Pike. (1995). How to use pheromone traps to monitor the Larger
Grain Borer (Prostephanus truncatus). Pamphlet produced by the Natural
Resources Institute, Central Avenue, Chatham Maritime, Kent ME4 4TB, UK.
ppl6. .
Holloway, G. J., and R. H. Smith. (1987). Sexual selection of body weight in Sitophilus
oryzae (L.) (Coleoptera: Curculionidae). Journal o f stored product research,
23:197-202.
Howard, D. C. (1983). The population biology of the greater grain borer Prostephanus
truncatus (Horn). PhD thesis, University of Reading.
Hughes, A. L. (1981). Differential male mating success in the white spotted sawyer
Monochamus scutellatus (Coleoptera: Cerambycidae). Annals of the entomological
soceity o f America., 74:180-184.
Ivarsson, P., F. Schlyter, and G. Birgersson. (1993). Demonstration of de novo

pheromone biosynthesis in Ips duplicatus (Coleoptera: Scolytidae) inhibition of
Ipsdienol and E-myrcenol production by compactin. Insect biochemistry and
molecular biology, 23:655-662.
Jaffe, K , P. Sanchez, H. Cerda, J. V. Hernandez, R. Jaffe, N. Urdaneta, G. Guerra, R.

M&rtinez, and B. Miras. (1993). Chemical ecology of the palm weevil
Rhynchophorus palmarium (L.) (Coleoptera: Curculionidae) - attraction to host
plants and to a male-produced aggregation pheromone. Journal of chemical
ecology, 19:1703-1720.
Karlson, P., and A. Butenandt. (1959). Pheromones (ectohormones) in insects. Annual

review o f entomology., 4:39-58.
Kerkut, G. A., and L. I. Gilbert. (1985). Comprehensive Insect Physiology. Biochemistry

and Pharmacology. Pergamon, New York.
Key, G. E., B. J. Tigar, E. Flores-Sanchez, and M. Vazquez-Arista. (1994). Response of

Prostephanus truncatus and Teretriosoma nigrescens to pheromone-baited flight
140
traps., 6th International Working Conference on Stored-product Protection., Vol.
1, pp. 410-414, Canberra, Australia.
Khorramshahi, A., and W. E. Burkholder. (1981). Behaviour of the lesser grain borer
Rhyzopertha dominica (Coleoptera: Bostrichidae) male-produced aggregation
pheromone attracts both sexes. Journal of chemical ecology, 7:33-38.
Kirkendall, L. R. (1983). The evolution of mating systems in bark and ambrosia beetles
(Coleoptera: Scolytidae and Platypodidae). Zoological journal o f the Linnean
society., 77:293-352.
Krausseopatz, B., U. Kohler, and R. Schopf. (1995). The energetic state of Ips
typographus L (Col, Scolytidae) during the life cycle. Journal o f applied
Landolt, P. J., and T. W. Phillips. (1997). Host plant influences on sex pheromone
behaviour of phytophagous insects. Annual Reviews o f Entomology, 42:371-391.
Lewis, S. M., and S. N. Austad. (1994). Sexual selection in flour beetles: the relationship
between sperm precedence and male olfactory attractiveness. Behavioural Ecology,
5:219-224.
Li, L. (1988). Behavioural ecology and Life history evolution in the Larger Grain Borer,
Prostephanus truncatus (Horn). Ph.D. thesis, Univeristy of Reading.
Marange, T., S. Floyd, and R. J. Hodges, (submitted). Technique for ageing adult
Prostephanus truncatus (Hom)(Coleoptera: Bostrichidae) studied using a novel
development chamber..
Markham, R. H., F. Djossou, J. M. Hirabayashi, P. Novillo, V. F. Wright, R. M. Rios,

F. J. Trujillo, W. G. Meikle, and C. Borgemeister. (1994). Biological control in
the context of an integrated management strategy for the Larger Grain Borer,
Prostephanus truncatus (Horn) (Coleoptera: Bostrichidae) and associated storage
pests., 6th International Working Conference on Stored-product Protection., Vol.
2, pp. 1106-1 111, Canberra, Australia.
Markham, R. H., V. F. Wright, and R. M. R. Ibarra. (1991). A selective review of

research on Prostephanus truncatus (Col., Bostrichidae) with an annotated and
updated bibliography. CEIBA, 32:1 -91.
Mayhew, T. J., and T. W. Phillips. (1994). Pheromone biology of the lesser grain borer,
Rhyzopertha dominica (Coleoptera: Bostrichidae), 6th International working
conference o f stored product protection., Vol. 1, pp. 541-544.
141
McCoy, J. R., and J. E. Wright. (1990). Selective breeding for increased pheromone
production in the Boll weevil (Coleoptera: Curculionidae). Journal of economic
Miller, D. R., and J. H. borden. (1992). (S)-(+)-Ipsdienol - Interspecific inhibition of Ips

latidens (LeConte) by Ips pini (Say) (Coleoptera, Scolytidae). Journal o f chemical
ecology., 18:1577-1582.
Miller, D. R., and J. H. Borden. (1992). (S)-(+)-Ipsdienol - Interspecific inhibition of Ips

latidens (LeConte) by Ips pini (Say) (Coleoptera: Scolytidae). Journal o f chemical
ecology., 18:1577-1582.
Miller, D. R., J. H. Borden, G. G. S. King, and K. N. Slessor. (1991). Ipsenol - an

aggregation pheromone for Ips latidens (LeConte) (Coleoptera: Scolytidae). Journal
Miller, D. R., J. H. Borden, and B. S. Lindgren. (1995). Verbenone - dose dependent

interruption of pheromone based attraction of 3 sympatric species of pine bark
beetles (Coleoptera, Scolytidae). Environmental entomology, 24:692-696.
Moore, A. J. (1988). Female preferences, male social status, and sexual selection in
Nauphoeta cinerea. Animal behaviour, 36:303-305.
Moore, A. J., N. L. Reagan, and K. F. Haynes. (1995). Conditional signalling strategies:

effects on ontogeny, social experience and social status on the pheromonal signal of
male cockroaches. Animal behaviour, 50:191-202.
Moore, P. J., N. L. Reagan-Wallin, K. F. Haynes, and A. J. Moore. (1997). Odour

conveys status on cockroaches. Nature, 389:25.
Nemec, V., V. Zumr, and P. Stary. (1993). Studies on the nutritional state and the
response to aggregation pheromones in the bark beetle, Ips typographus (L.) (Col:
Scolytidae). Journal o f applied entomology, 116:358-363.
Nielsen, B. S., and T. S. Jensen. (1993). Spring dispersal of Sitona lineatus. The use of
aggregation pheromone traps for monitoring. Entomologia experimentalis et
applicata, 66:21-30.
Nijholt, W. W. (1973). The effect of male Trypodendron lineatum on the response of field
populations to secondary attraction. Canadian Entomol., 105:583-590.
Obeng-Ofori, D., and T. H. Coaker. (1990). Some factors affecting responses of 4 stored
product beetles (Coleoptera: Tenebrionidae and Bostrichidae) to pheromones.
Bulletin o f entomological research, 80:433-441.
142
Paine, T. D., and C. C. Hanlon. (1991). Response of Dendroctonus brevicomis and Ips
paraconfusus (Coleoptera, Scolytidae) to combinations of synthetic pheromone
attractants and inhibitors verbenone and ipsdienol. Journal o f chemical ecology.,
17:2163-2176.
Parker, G. A. (1970). Sperm competition and it's evolutionary consequences in the insects.
Biological review., 45:525-567.
Patrock, R. J., D. J. Schuster, and E. R. Mitchell. (1992). Field evidence for an attractant
produced by the male pepper weevil (Coleoptera: Curculionidae). Florida
entomologist., 75:138-144.
Payne, T. L., R. F. Billings, C. W. Berisford, S. M. Salom, D. M. Grosman, M. J.

Dalusky, and W. W. Upton. (1992). Disruption of Dendroctonus frontalis (Col.,
Scolytidae) infestations with an inhibitor pheromone. Journal o f applied
entomology., 114:341-347.
Peng, C. W., and M. J. Weiss. (1992). Evidence of an aggregation pheromone in the flea
beetle, Phyllotreta cruciferae (Goese) (Coleoptera: Chrysomelidae). Journal of
applied entomology, 18:875-884.
Perez, A. L., G. Gries, R. Gries, R. M. Giblindavis, and A. C. Oehlschlager. (1994).

Pheromone chirality of African palm weevil, Rhynchophorus phoenicis (F.) and
palmetto weevil, Rynchophorus cruentatus (F.) (Coleoptera, Curculionidae).
Journal o f chemical ecology., 20:2653-2671.
Petrie, M. (1992). Copulation behaviour in birds: Why do females copulate more than once
with the same male? Animal behaviour, 44:790-792.
Petroski, R. J., R. J. Bartelt, and R. S. Vetter. (1994). Male produced aggregation

pheromone of Carpophilus obsoletus (Coleoptera: Nitidulidae). Journal o f chemical
ecology, 20:1483-1493.
Petroski, R. J., and R. Vaz. (1995). Insect aggregation pheromone response synergized by
host type volatiles. Molecular modeling evidence for close proximity binding of
pheromone and coattractant in Carpophilus hemipterus (L.) (Coleoptera,
Nitidulidae). ACS Symposium series, 589:197-210.
Phillips, T. W. (1994). Pheromones of stored-product insects: current status and future

perspectives., 6th International Working Conference on Stored-product Protection.,
Vol. 1, pp. 479-486, Canberra, Australia.
Phillips, T. W., X. L. Jiang, W. E. Burkholder, J. K. Phillips, and H. Q. Tran. (1993).

Behavioural responses to food volatiles by 2 species of stored product Coleoptera,
143
Sitophilus oryzae (Curculionidae) and Tribolium castaneum (Tenebrionidae).
Journal o f chemical ecology, 19:723-734.
Pierce, A. M., H. D. Pierce, A. C. Oehlschlager, and J. H. Borden. (1991). l-octen-3-ol,

attractive semiochemical for foreign grain beetle, Ahasverus advena (waltl.)
(Coleoptera, Cucujidae). Journal o f chemical ecology, 17:567-580.
Pierce, H. D., P. Degroot, J. H. Borden, S. Ramaswamy, and A. C. Oehlschlager.

(1995). Pheromones in red pine cone beetle, Conophorus resinosae Hopkins, and
its synonym, C. Banksianae McPherson (Coleoptera: Scolytidae). Journal of
chemical ecology., 21:169-185.
Pike, V. (1993). Development of pheromone-baited traps and their uses for the Larger
Grain Borer. Bulletin o f the International organisaionfor the biological and
integrated control o f noxious animals and plants, 16:64-70.
Pike, V., J. L. Smith, R. D. White, and D. R. Hall. (1994). Studies of responses of

stored-product pests, Prostephanus truncatus (Horn) and Sitophilus zeamais
Motsch., to food volatiles., 6th International Working Conference on Stored-
product Protection., Vol. 1, pp. 566-569, Canberra, Australia.
Pitman, G. B. (1969). Pheromone response in Pine bark beetles: Influence of host

volatiles. Science, 166:905-906.
Prestwich, G. D., and G. J. Blomquist. (1987). Pheromone biochemistry. Academic

press, London.
Raffa, K. F., and D. L. Dahlsten. (1995). Differential responses among natural enemies
and prey to bark beetle pheromones. Oecologia, 102:17-23.
Raffa, K. F., T. W. Phillips, and S. M. Salom. (1993). Strategies and mechanisms of host
colonization by bark beetles. In R. D. Schowalter and G. M. Filip (eds.), Beetle-
pathogen interactions in conifer forests., pp. 103-127. Academic, New York.
Raina, A. K., and E. A. Stadelbacher. (1990). Pheromone titer and calling in Heliothis
virescens (Lepidoptera: Noctuidae): Effect of mating with normal and sterile
backcross males. Annuls o f the entomological society o f America., 83:987-990.
Raina, A. K., and E. A. Stadelbacher. (1990). Pheromone titer and calling in Heliothis
virescens (Lepidoptera: Noctuidae): Effect of mating with normal and sterile
backcross males. Annuls o f the entomological society o f America., 83:987-990.
144
Rangaswamy, J. R., and V. B. Sasikala. (1991). Aggregation pheromone activity of
compounds isolated from male red four beetle Tribolium castaneum (Coleoptera:
Tenebrionidae). Indian journal of experimental biology, 29:52-55.
Renwick, J. A. A., and J. P. Vite. (1969). Bark beetle attractants: mechanism of

colonisation by Dendroctonus frontalis. Nature., 224:1222-1223.
Ridley, M. (1989). The incidence of sperm displacement in insects: four conjectures, one
corroboration. Biological journal o f the Linnean society, 38:349-367.
Rochat, D., C. Descoins, C. Malosse, P. Nagnan, P. Zagatti, F. Akamou, and D. Mariau.

(1993). Chemical ecology of palm weevils, Rhynchophorus spp. (Coleoptera).
Oleagineux, 48:225-236.
Rochat, D., A. Gonzalez, D. Mariau, A. Villanueva, and P. Zagatti. (1991a). Evidence for
male produced aggregation pheromone in American palm weevil, Rhynchophorus
palmarum (L.) (Coleoptera: Curculionidae). Journal o f chemical ecology.,
17:1221-1230.
Rochat, D., C. Malosse, M. Lettere, P. H. Ducrot, P. Zagatti, M. Renou, and C.

Descoins. (1991b). Male produced aggregation pheromone of the American palm
weevil, Rhynchophorus palmarium (L.) (Coleoptera: Curculionidae). Collection,
identification, electrophysiological activity, and laboratory bioassay. Journal o f
chemical ecology., 17:2127-2141.
Ross, D. W., and G. E. Daterman. (1994). Reduction of Douglas fir beetle infestations of
high risk stands by antiaggregation and aggregation pheromones. Canadian journal
o f forest research., 24:2184-2190.
Rudinsky, J. A. (1963). Response of Dendroctonus pseudotsuage Hopkins to volatile

attractants. Contr. Boyce Thompson Inst. PI. Res., 22:22-38.
Safranyik, L., T. L. Shore, D. A. Linton, and B. S. Lindren. (1992). The effect of

verbenone on dispersal and attack of the mountain pine beetle, Dendroctonus
ponderosae Hopk (Col, Scolytidae) in a lodgepole pine stand. Journal of applied
entomology., 113:391-397.
Salom, S. M., R. F. Billings, W. W. Upton, M. J. Daluski, D. M. Grosman, T. L.

Payne, C. W. Berisford, and T. N. Shaver. (1992). Effect of verbenone
enantiomers and racemic endo-brevicomin on response of Dendroctonus frontalis
(Coleoptera: Scolytidae) to attractant baited traps. Canadian journal offorest
research., 22:925-931.
145
Schlyter, F., and O. Anderbrant. (1993). Competition and niche seperation between 2 bark
beetles, existence and mechanisms. Oikos, 68:437-447.
Schlyter, F., and G. Birgersson. (1989). Individual variation in bark beetle and moth
pheromones a comparison and an evolutionary background. Holarc. Ecology,
12:457-465.
Schlyter, F., J. A. Byers, and J. Loefqvist. (1987). Attraction to pheromone sources of

different quantity, quality, and spacing. Density regulation mechanisms in bark
beetle Ips typographies. Journal o f chemical ecology, 13:1503-1523.
Schmitz, R. F. (1972). Behaviour of Ips pini during mating, oviposition, and larval
development (Coleoptera: Scolytidae). The Canadian entomologist, 104:1723-
1728.
Scholz, D. (1997). Dispersal and host finding behaviour of Prostephanus truncatus (Horn)
(Coleoptera: Bostrichidae)., Hannover University.
Scholz, D., C. Borgemeister, W. E. Meikle, R. H. Markham, and M. P. H. (1997a).

Infestation of maize by Prostephanus truncatus initiated by male-produced
pheromone. Entomologia experimental et applicata, 83:
Scholz, D., A. Tchabi, C. Borgemeister, R. H. Markham, H. M. Poehling, and A.

Lawson. (1997b). Host finding behaviour of Prostephanus truncatus
(Hom)(Col.,Bostrichidae): primary attraction or random attack? Journal of applied
Seybold, S. J. (1993). Role of chirality in olfactory directed behaviour - aggregation of

pine engraver beetles in the genus Ips (Coleoptera: Scolytidae). Journal o f chemical
ecology., 19:1809-1831.
Seybold, S. J., D. R. Quilici, J. A. Tillman, D. Vanderwel, D. L. Wood, and G. J.

Blomquist. (1995). De-novo biosynthesis of the aggregation pheromone
components Ipsenol and Ipsdienol by the pine bark beetles Ips paraconfusus
(Lanier) and Ips pini (Say) (Coleoptera: Scolytidae). Proceedings of the national
academy o f sciences o f the United States o f America., 92:8393-8397.
Shires, S. W. (1980). Life history of Prostephanus trunctus (Horn) (Coleoptera:

Bostrichidae) at optimum conditions of temperature and humidity. Journal o f stored
products research, 16:147-150.
Shires, S. W., and S. McCarthy. (1976). A character for sexing live adults of
Prostephanus truncatus (Horn) (Bostrichidae: Coleoptera). Journal stored products
research., 12:273-275.
146
Shore, T. L., L. Safranyik, and B. S. Lindgren. (1992). The response of mountain pine
beetle (Dendroctonus ponderosae) to lodgepole pine trees baited with verbenone and
exo-brevicomin. Journal o f chemical ecology., 18:533-541.
Shorey, H. H. (1973). Behavioural responses to insect pheromones. Annual review of

Entomology, 18:349-380.
Silverstein, R. M. (1990). Practical use of pheromones and other behaviour modifying

compounds: Overview. In R. L. Ridgway, R. M. Silverstein and M. N. Inscoe
(eds.), Behaviour modifying chemicals for insect management. Application of
pheromones and other attractants. Marcel Dekker, Inc., New York.
Simmons, L. W., and G. A. Parker. (1992). Individual variation in sperm competition

success of yellow dung flies, Scatophaga sterroraria. Evolution, 46:366-375.
Smart, L. E., M. M. Blight, J. A. Pickett, and B. J. Pye. (1994). Development of field

strategies incorporating semiochemicals for the control of the pea and bean weevil,
Sitona lineatus L. Crop protection, 13:127-135.
Smith, J. L., A. Cork, D. R. Hall, and R. J. Hodges. (1996). Investigation of the effect of
female Larger Grain Borer, Prostephanus truncatus (Horn) (Coleoptera:
Bostrichidae), and their residues on the production aggregation pheromone by
males. Journal of stored products research, 32:171-181.
Sokal, R. R., and F. J. Rohlf. (1981). Biometry. 2nd Edition ed. W.H. Freeman and
company, San Francisco.
Svensson, M. (1996). Sexual selection in moths: The role of chemical communication.

Biological Review, 71:113-135.
Teale, S. A., B. J. Hager, and F. X. Webster. (1994). Pheromone based assortative

mating in a bark beetle. Animal behaviour, 48:569-578.
Teale, S. A., F. X. Webster, A. J. Zhang, and G. N. Lanier. (1991a). Lanierone - a new

pheromone component from Ips pini (Coleoptera: Scolytidae) in New York. Journal
Thornhill, R., and J. Alcock. (1983). The evolution of insect mating systems. Harvard
University press, Cambridge MA.
Thornhill, R., and J. Alcock. (1983). The evolution of insect mating systems. Harvard
University press, Cambridge MA.
147
Tigar, B. J., G. E. Key, M. E. Flores, and M. Vazquez. (1993). Flight periodicity of
Prostephanus truncatus and longevity of attraction to synthetic pheromone.
Entomologia experimentalis et applicata, 66:91-97.
Vanderwel, D., B. Johnston, and A. C. Oehlschlager. (1992). Cucujolide biosynthesis in

the Merchant and Rusty grain beetles. Insect biochemistry and molecular biology.,
22:875-883.
Vanderwel, D., and A. C. Oehlschlager. (1987). Chapter 6. Biosynthesis of pheromones

and endocrine regulation of pheromone production in Coleoptera. In G. D.
Prestwich and G. J. Blomquist (eds.), Pheromone biochemistry, pp. 175-215.
Academic press, London.
Vazquez-Arista, M. (1997). Anatomical, enzymatic, and microbiological studies on the

digestive system of Prostephanus truncatus (Horn). Ph.D. thesis, University of
Leicester.
Visser, J. H. (1986). Host odor perception in phytophagous insects. Annul Review of

Entomology., 31:121-144.
Vite, J. P., and G. B. Pitman. (1968). Bark beetle aggregation: Effects of feeding on the
release of pheromones in Dendroctonus and Ips. Nature, 218:169-170.
Walgenbach, C. A., and W. E. Burkholder. (1986). Factors affecting the response of the
maize weevil, Sitophilus zeamais (Coleoptera: Curculionidae), to its aggregation
pheromone. Environmental entomology, 15:733-738.
Walgenbach, C. A., J. K. Phillips, and D. L. Faustini. (1983). Male produced aggregation

pheromone of the maize weevil Sitophilus zeamais and interspecific attraction
between 3 Sitophilus species. Journal of chemical ecology, 9:831-841.
Weissling, T. J., R. M. Giblin-Davis, G. Gries, R. Gries, A. L. Perez, H. D. P. Jr., and

A. C. Oehlschlager. (1994). Aggregation pheromone of Palmetto weevil,
Rhynchophorus cruentatus (F.) (Coleoptera: Curculionidae). Journal o f chemical
ecology., 20:505-515.
Wekesa, P. W. (1994). Field and store ecology of the Larger Grain Borer, Prostephanus
truncatus (Horn) (Coleoptera: Bostrichidae) in Kenya. Ph.D. thesis, Univeristy of
Leicester.
White, P. R., and J. Chambers. (1989). Saw toothed grain beetle Oryzaephilus
surinamenis (L.) (Coleoptera: Silvanidae). Antennal and behavioural responses to
individual components and blends of aggregation pheromone. Journal o f chemical
ecology, 15:1015-1031.
148
White, P. R., J. Chambers, C. M. Walter, J. P. G. Wilkins, and J. G. Millar. (1989).
Saw toothed grain beetle Oryzaephilus surinamensis (L.) (Coleoptera: Silvanidae).
Collection, identification, and bioassay of attractive volatiles from beetles and oats.
Journal o f chemical ecology, 15:999-1013.
Wigglesworth, V. B. (1972). The principles of insect physiology. 7th ed. Chapman and
Hall, London.
Wilkinson, L. (1990). SYSTAT: The system for statistics. SYSTAT press, Evanston,
Illinois.
Wirtz, P. (1997). Sperm selection by females. TREE, 12:172-173.
Wood, D. L. (1982). The role of pheromones, kairomones, and allomones in the host
selection and colonization behaviour of bark beetles. Annual review o f entomology,
27:411-446.
Wood, D. L., L. E. Browne, W. D. Bedard, P. E. Tilden, R. M. Silverstein, and O. J.

Rodin. (1968). Response of Ips confusus to synthetic sex pheromones in nature.
Science, 159:1373-1374.
Zeh, A. J., and D. W. Zeh. (1994). Last-male sperm precedence breads down when
females mate with three males. Proceedings of the Royal Society o f London. B,
257:287-292.
Zuber, M., and G. Benz. (1992). Investigations on the temporary dequence of the flight
activities of Ips typographus (L.) and Pitogenes chalcographus (L.) (Col.,
Scolytidae) by means of the commercial pheromone preparations Pheroprax and
Chalcoprax. Journal o f applied entomology., 113:430-436.
Zuber, M., H. Meyer, U. Kohnle, and W. Francke. (1993). Odour production and
pheromone response in the European engraver bark beetles, Ips amitinus var.
montana Fuchs (Col: Scolytidae). Journal o f applied entomology., 115:462-465.
149

Birkinshaw 1998

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Birkinshaw 1998

Uploaded by

Copyright:

Available Formats

Mate choice in Pros

Thesis submitted for the degree of

All rights reserved

INFORMATION TO ALL USERS

I thank my supervisor, Professor Robert ‘thrasher’ Smith, for a great three

Sterilization of insects was made possible by the kind permission and

I thank my aged ancestors, damned average genetic inheritance! Life in

Lucy A. Birkinshaw 1998: Mate choice in Prostephanus truncatus (Horn)

Prostephanus truncatus (Horn) (Coleoptera: Bostrichidae) is a destructive pest

The literature on Coleopteran aggregation pheromones was reviewed, with

Variation in Prostephanus truncatus aggregation-pheromone signalling was

Observation of adult beetles in an artificial host sandwiched between two glass

CHAPTER 1: INTRO DUCTIO N............................................................................1

1.1 THE INSECT..............................................................................................................................1

1.2 STARTING HYPOTHESIS.....................................................................................................4

1.3 SEXUAL SELECTION............................................................................................................ 5

1.4 SUMMARY OF CHAPTERS................................................................................................ 6

1.5 GENERAL MATERIALS ANDMETHODS.........................................................................7

CHAPTER 2: LITERATURE REVIEW OF COLEOPTERAN

2.5: ANTI-AGGREGATION PHEROMONES........................................................................ 18

2.6: NON ADAPTIVE EXPLANATIONS................................................................................ 19

2.7: ADAPTIVE EXPLANATIONS........................................................................................... 19

2.8: COSTS OF SIGNALLING.................................................................................................20

2.9: BENEFITS OF SIGNALLING...........................................................................................20

2.10: COSTS AND BENEFITS OF RESPONDING TO PHEROMONE.........................23

2.11: HYPOTHESES DISCUSSED IN TERMS OF THE BIOLOGICAL FEATURES

2.12: FUTURE DIRECTIONS.................................................................................................... 26

CHAPTER 3: BASIC PATTERNS OF AGGREGATION-PHEROMONE

3.2 BIOASSAY DESIGN............................................................................................................. 3 0

3.3 MATERIALS AND M ETHODS......................................................................................... 33

CHAPTER 4: INTER-MALE VARIATION IN PHEROMONE SIGNAL (I

4.3 MATERIALS AND METHODS 47

CHAPTER 5: INTER-MALE VARIATION IN PHEROMONE SIGNAL (III:

5.3 MATERIALS AND METHODS........................................................................................ 65

CHAPTER 6: MATE CHOICE ON CONTACT............................................. 7 9

6.3 DESCRIPTION OF COURTSHIP BEHAVIOUR IN PR O STEPH AN U S

6.4 MATERIALS AND METHODS........................................................................................ 83

CHAPTER 7: SPERM C O M P E TITIO N ........................................................ 1 0 5

7.1 INTRODUCTION................................................................................................................ 105

7.2 A IM S...................................................................................................................................... 107

7.3 METHODS............................................................................................................................. 108

7.5 DISCUSSION....................................................................................................................... 122

CHAPTER 8: DISCU SSIO N............................................................................1 2 7

8.1 Evolution of the aggregationpheromone signal/ response:...................................... 127

8.2 Alternative mating tactics:..............................................................................................131

8.3 The limitations of an evolutionaryapproach:.............................................................. 132

8.4 Practical applications of this work:.................................................................................132

8.5 Suggestions for future work:...........................................................................................133

Fig. 1: Adult Prostephanus truncatus (length 2.5-4.5mm), and larva

1.1 THE INSECT

1.1.2 Life history

If a suitable host is available a female can consistently lay an average of four

1.1.3 Host range

1.1.4 Research em phasis to date

Two components of LGB aggregation pheromone have been identified and

Host-plant volatiles are often used to increase the efficiency of aggregation-

1.2 STARTING HYPOTHESIS

1.3 SEXUAL SELECTION

Two of the earliest mechanisms proposed to result in sexual selection were

1.3.1 Low sexual dimorphism of LGB external morphology

1.4 SUMMARY OF CHAPTERS

1.5 GENERAL MATERIALS AND METHODS

1.5.2 Source of plant material used as insect hosts

1.5.3 Insect cultures